Frontostriatal salience network expansion in individuals in depression

Abstract

Decades of neuroimaging studies have shown modest differences in brain structure and connectivity in depression, hindering mechanistic insights or the identification of risk factors for disease onset¹. Furthermore, whereas depression is episodic, few longitudinal neuroimaging studies exist, limiting understanding of mechanisms that drive mood-state transitions. The emerging field of precision functional mapping has used densely sampled longitudinal neuroimaging data to show behaviourally meaningful differences in brain network topography and connectivity between and in healthy individuals2–4, but this approach has not been applied in depression. Here, using precision functional mapping and several samples of deeply sampled individuals, we found that the frontostriatal salience network is expanded nearly twofold in the cortex of most individuals with depression. This effect was replicable in several samples and caused primarily by network border shifts, with three distinct modes of encroachment occurring in different individuals. Salience network expansion was stable over time, unaffected by mood state and detectable in children before the onset of depression later in adolescence. Longitudinal analyses of individuals scanned up to 62 times over 1.5 years identified connectivity changes in frontostriatal circuits that tracked fluctuations in specific symptoms and predicted future anhedonia symptoms. Together, these findings identify a trait-like brain network topology that may confer risk for depression and mood-state-dependent connectivity changes in frontostriatal circuits that predict the emergence and remission of depressive symptoms over time.

Full text

624 | Nature | Vol 633 | 19 September 2024
Article
Frontostriatal salience network expansion in
individuals in depression
Charles J. Lynch1 ✉, Immanuel G. Elbau1, Tommy Ng1, Aliza Ayaz1, Shasha Zhu1, Danielle Wolk1,
Nicola Manfredi1, Megan Johnson1, Megan Chang1, Jolin Chou1, Indira Summerville1,
Claire Ho1, Maximilian Lueckel2,3, Hussain Bukhari1, Derrick Buchanan4, Lindsay W. Victoria1,
Nili Solomonov1, Eric Goldwaser1, Stefano Moia5,6,7, Cesar Caballero-Gaudes7,
Jonathan Downar8, Fidel Vila-Rodriguez9, Zairis J. Daskalakis10, Daniel M. Blumberger8,11,12,
Kendrick Kay13, Amy Aloysi14, Evan M. Gordon15, Mahendra T. Bhati4, Nolan Williams4,
Jonathan D. Power1, Benjamin Zebley1, Logan Grosenick1, Faith M. Gunning1 & Conor Liston1 ✉
Decades of neuroimaging studies have shown modest dierences in brain structure
and connectivity in depression, hindering mechanistic insights or the identication
of risk factors for disease onset1. Furthermore, whereas depression is episodic, few
longitudinal neuroimaging studies exist, limiting understanding of mechanisms that
drive mood-state transitions. The emerging eld of precision functional mapping
has used densely sampled longitudinal neuroimaging data to show behaviourally
meaningful dierences in brain network topography and connectivity between and
in healthy individuals2–4, but this approach has not been applied in depression.
Here, using precision functional mapping and several samples of deeply sampled
individuals, we found that the frontostriatal salience network is expanded nearly
twofold in the cortex of most individuals with depression. This eect was replicable
in several samples and caused primarily by network border shifts, with three distinct
modes of encroachment occurring in dierent individuals. Salience network
expansion was stable over time, unaected by mood state and detectable in children
before the onset of depression later in adolescence. Longitudinal analyses of
individuals scanned up to 62 times over 1.5 years identied connectivity changes in
frontostriatal circuits that tracked uctuations in specic symptoms and predicted
future anhedonia symptoms. Together, these ndings identify a trait-like brain
network topology that may confer risk for depression and mood-state-dependent
connectivity changes in frontostriatal circuits that predict the emergence and
remission of depressive symptoms over time.
Depression is a heterogeneous and episodic neuropsychiatric syn-
drome associated with synapse loss
5,6
and connectivity alterations
in frontostriatal networks
7–9
and a leading cause of disability world-
wide10. The neurobiological mechanisms that give rise to specific
depressive symptom domains or to changes in mood over time are
not well understood, especially at the neural systems level. So far,
most functional magnetic resonance imaging (fMRI) studies have
tested for differences in functional connectivity in cross-sectional
comparisons between groups of depressed individuals and healthy,
never-depressed controls using group-average (one-size-fits-all)
parcellations to define functional brain areas and networks. More
recently, pioneering work in systems neuroscience has given rise to
the field of precision functional mapping, which refers to a suite of
new approaches for delineating functional networks entirely in indi-
viduals
2–4,11–14
. Precision mapping studies have shown that the topology
(size, shape and spatial location) of functional areas and networks in
individuals deviates markedly from group-average descriptions3,15,16 and
that individual differences in network topology are stable2,3,17,18, herit-
able19,20 and associated with cognitive abilities and behaviour3–5,13,21–23.
Apart from a recent case study involving a single individual who sus-
tained bilateral perinatal strokes
24
, these tools have not yet been widely
applied in clinical populations, including depression. Thus, whether
https://doi.org/10.1038/s41586-024-07805-2
Received: 2 August 2023
Accepted: 9 July 2024
Published online: 4 September 2024
Open access
Check for updates
1Department of Psychiatry, Weill Cornell Medicine, New York, NY, USA. 2Leibniz Institute for Resilience Research, Mainz, Germany. 3Neuroimaging Center (NIC), Focus Program Translational
Neurosciences (FTN), Johannes Gutenberg University Medical Center, Mainz, Germany. 4Department of Psychiatry and Behavioral Sciences, Stanford University, Stanford, CA, USA.
5Neuro-X Institute, École Polytechnique Fédérale de Lausanne, Geneva, Switzerland. 6Department of Radiology and Medical Informatics, Faculty of Medicine, University of Geneva, Geneva,
Switzerland. 7Basque Center on Cognition, Brain and Language, Donostia, Spain. 8Department of Psychiatry and Institute of Medical Science, University of Toronto, Toronto, Ontario, Canada.
9Department of Psychiatry, University of British Columbia, Vancouver, British Columbia, Canada. 10Department of Psychiatry, University of California, San Diego, CA, USA. 11Temerty Centre
for Therapeutic Brain Intervention, Toronto, Ontario, Canada. 12Centre for Addiction and Mental Health, Toronto, Ontario, Canada. 13Center for Magnetic Resonance Research, University of
Minnesota, Minneapolis, MN, USA. 14Icahn School of Medicine at Mount Sinai, New York, NY, USA. 15Department of Radiology, Washington University School of Medicine, St. Louis, MO, USA.
✉e-mail: cjl2007@med.cornell.edu; col2004@med.cornell.edu
Content courtesy of Springer Nature, terms of use apply. Rights reserved

Nature | Vol 633 | 19 September 2024 | 625
functional network topology differs in individuals with depression
is unknown.
Depression is a fundamentally episodic neuropsychiatric condition
defined by discrete periods of low mood interposed between periods
of euthymia but our understanding of the mechanisms that mediate
mood transitions over time is limited. This is due in part to the fact
that most studies until now have been cross-sectional, involving data
acquired at a single time point or, in some cases, two or three scans
acquired before and after an intervention25–27—an approach that is
not designed for meaningful statistical inferences at the individual
level21. Understanding the neurobiological mechanisms that mediate
transitions in and out of depressive mood states may require studying
individual patients over many months
22
. Indeed, densely sampled n-of-1
studies involving intracranial EEG recordings and other assessments
have begun to show mechanisms that regulate mood-state transitions
in individual patients receiving deep brain stimulation for depres-
sion23,28,29 but these approaches have not yet been deployed at scale in
fMRI studies. Without such datasets, it is unknown whether changes
in brain network connectivity predict the emergence of anhedonia,
anxiety and dysfunction in other depressive symptom domains or
the subsequent remission of these symptoms after a recovery from an
episode. In the same way, it is unclear whether atypical network topol-
ogy measures fluctuate with mood state in individuals with depres-
sion or remain stable over time—key questions for understanding
cause-and-effect relationships in clinical neuroimaging, for defining
potential therapeutic targets in neuromodulation interventions or
identifying at-risk individuals. Until recently, technical limitations
have posed significant obstacles to performing precision functional
mapping and longitudinal neuroimaging in clinical samples, including
depression. Conventional fMRI measurements at the single subject level
are often noisy and have limited reliability, in part because they are sen-
sitive to a variety of imaging artefacts30. However, recent studies have
taken significant steps towards developing solutions to these problems,
by either acquiring large quantities of data in each subject
2,3,14
or by
using multi-echo fMRI
18,31
. Together, these approaches can generate
highly reliable functional connectivity measures and network maps at
the level of individual subjects, an important step towards developing
and deploying fMRI for clinical translational purposes.
Here we used state-of-the-art precision functional mapping tools
to delineate topology of functional brain networks in individuals with
depression, leveraging several resting-state fMRI datasets of deeply
sampled individuals. We found that the frontostriatal salience network
is expanded by nearly twofold in most individuals with depression—an
effect we replicated thrice using independent samples of repeatedly
sampled individuals with depression (total n = 135) and in large-scale
group-average data (n = 299 individuals with depression, n = 932 healthy
controls), with three distinct types of encroachment displacing neigh-
bouring functional systems occurring across individuals. Salience
network expansion was stable over time and unaffected by changes in
mood state. It was also present in children scanned before the onset of
depression symptoms that emerged later in adolescence. Longitudinal
analyses of densely sampled individuals showed mood-state-dependent
changes in striatal connectivity with anterior cingulate and anterior
insular nodes of the salience network which tracked fluctuations in
anhedonia and anxiety, respectively, and predicted the subsequent
emergence of anhedonic symptoms at future study visits.
Salience network expansion in depression
Numerous neuroimaging studies involving large cohorts of patients
with depression have identified differences in functional connectivity
and brain structure
32–36
, often involving the anterior cingulate cortex,
orbitofrontal cortex, insular cortex and subgenual cingulate cortex—a
therapeutic target for deep brain stimulation
37,38
—but the effect sizes in
large-scale meta-analyses are modest (for example, Cohen’s d = 0.1–0.15
for structural measures
36
and d = 0.13–0.26 for functional connectivity
measures
1
). Whether the topological features of large-scale functional
brain networks—their shape, spatial location and size—are altered in
depression is unknown.
We used precision functional mapping to delineate the topology of
functional brain networks in six highly sampled individuals with unipolar
major depression who underwent on average 621.5 min of multi-echo
fMRI scanning (range 58–1,792 min) across 22 sessions (range 2–62
sessions). We refer to this dataset as the serial imaging of major depres-
sion (SIMD) dataset (study design and aims in Extended Data Fig.1). To
contextualize the severity of depressive symptoms in these individuals,
the mean 17-item
39
Hamilton depression rating scale (HDRS17) score
(averaged across study visits, excluding those when these individuals
were in remission) was 15.7 ± 3.7 (range 10.5–22.2), indicating a range of
severity levels from mild to severe. The same precision mapping proce-
dures were applied to 37 highly sampled healthy controls with an average
of 327.49 min of fMRI data per subject (range 43.36–841.2 min) across 12
sessions (range 2–84 sessions). The healthy controls did not receive any
intervention or treatment. See theMethods for more details.
It was immediately apparent on visual inspection that the salience
network, which is involved in reward processing and conscious inte-
gration of autonomic feedback and responses with internal goals and
environmental demands
30,40,41
, was markedly larger in these individuals
with depression (Fig.1a,b). In four of the six individuals, the salience
network was expanded more than twofold, outside the range observed
in all 37 healthy controls (Fig.1c, left). On average, the salience net-
work occupied 73% more of the cortical surface relative to the mean in
healthy controls (5.49% ± 0.76% of cortex in SIMD versus 3.17% ± 0.85%
of cortex in healthy controls), giving rise to a large group-level effect
(Cohen’s d = 1.99). This effect was replicated using an alternative net-
work parcellation algorithm42 (Supplementary Fig.1) and without use
of global signal regression (Supplementary Fig.2), indicating that it is
robust to methodological variation and was not explained by group
differences in brain anatomy or structure (Supplementary Fig.3) or
head motion (independent sample t-test comparing mean framewise
displacement, T = 0.73, P = 0.47, claims about equivalence are based
on an absence of evidence).
To further validate this finding, we repeated this procedure in three
samples (n = 48 and n = 45 from Weill Cornell Medicine and n = 42 from
Stanford University) of individuals with depression. Detailed imaging,
demographic and clinical information for these samples are available
in Supplementary Table1 and Supplementary Fig.4. The effect was
replicated thrice (Fig.1c, right), again with medium to large effect sizes
(Cohen’s d = 0.77–0.84), remained statistically significant when control-
ling for the sex ratio imbalance in our samples (56.7% of individuals with
depression were female, versus 31% of the healthy controls; Supplemen-
tary Fig.4 and Extended Data Fig.2) and with or without correction for
potential site- or scanner-induced biases (Supplementary Fig.5). We
also evaluated if representation of the salience network was similarly
increased in the striatum, which is thought to relate to anatomically
well-defined, interconnected loops in which the cortex projects to the
striatum and the striatum projects back to cortex indirectly through
the thalamus
43,44
but found that the difference in group means was not
statistically significant (Fig.1d).
Salience network expansion in depression was also evident in density
maps (Fig.1e), which convey the percentage of individuals with salience
network representation at each cortical vertex or striatal voxel. These
maps confirmed a similar overall pattern of cortical and subcortical
representation in both groups, consistent with descriptions in previous
reports3,45,46 but also showed that the borders of the salience network
frequently extended further outwards from their centroids in each
cortical zone in depressed individuals. For example, in the anterior
cingulate cortex, network borders shifted more anteriorly into the
pregenual cortex and, in lateral prefrontal cortex, network borders
shifted more anteriorly towards the frontal pole (red boxes in Fig.1e).
Content courtesy of Springer Nature, terms of use apply. Rights reserved

626 | Nature | Vol 633 | 19 September 2024
Article
Accordingly, expansion of the salience network in cortex was accom-
panied by contraction of neighbouring functional systems in the
SIMD sample (Extended Data Fig.3). However, the specific patterns
of contraction did not replicate in all datasets—a finding we return
to in the following section. Otherwise, consistent and reproducible
group differences in network size were specific to the salience network
(no significant differences in the size of any other network after cor-
recting for several comparisons).
To better understand whether this effect was also detectable in large,
previously published samples involving conventional single-echo fMRI
data, we identified the salience network in group-average functional
connectivity data from two large datasets involving n = 812 (ref. 47) and
n = 120 (ref. 45) healthy controls, respectively, and in a third dataset
involving n = 299 individuals with treatment-resistant depression
scanned in association with a neuromodulation intervention study48.
The cortical representation of the salience network was more than 70%
larger in the 299-subject depression sample compared to two healthy
control samples (Extended Data Fig.4). Furthermore, highly similar
patterns of salience network topology and functional connectivity were
produced in split-half analyses of each SIMD dataset (Extended Data
Fig.5), indicating that salience network expansion was a robust and
reproducible feature of the brains of these highly sampled individuals.
Given the magnitude of the effect reported in Fig.1c, we went on to
test whether individuals with depression could be distinguished algo-
rithmically from healthy controls using only the size of each functional
network as predictive features. Thus, we trained a linear support vector
HC
MDD-1
MDD-2
MDD-3
0
2
4
6
8
10
Percentage
of cortex
*
SAL in healthy controls
(3.17% of cortex, on average)
SIMD-1
(5.71% of cortex)
ab
Striatum
Percentage of individuals
>20 >80
PU
NAc
d Discovery Replications
SVM (nested split-half cross-validation)
AI ACC
LPFC
SIMD-2
(4.61% of cortex)
SIMD-4
(5.27% of cortex)
e
Healthy controls (n = 37) Depression (n = 141)
Cd
f
AI ACC
LPFC
Cortical surface
gLeave one network out
SAL density maps
HC
SIMD
0
2
4
6
8
10
ReplicationsDiscovery
c
*
HC
SIMD
0
20
40
60
80
100
HC
MDD-1
MDD-2
MDD-3
0
20
40
60
80
100
Percentage
of striatum
0
25
50
75
100
Accuracy (%)
Confusion matrix100
62.1% 37.9%
Subjects (%)
Predicted
17.5% 82.5%
HC MDD 0
Null
distribution
Feature weights
Average
accuracy = 78.4%
HC
Group
MDD
h
0
0.2
0.4
0.6
0.8
1.0
Weights |E|
SAL
Other
networks
–10
–8
–6
–4
–2
0
Accuracy loss (%)
Other
networks
SAL
Default-
parietal
Percentage
of striatum
Percentage
of cortex
Fig. 1 | Frontos triatal sa lience net work is expanded n early twofold in th e
cortex of highly sampled individuals with depression. a, The salience
network (b lack) has repres entation in LP FC, ACC and AI. b, The sa lience
network in th ree represent ative individu als from the data set referred to
here as ser ial imaging of maj or depression (SI MD).c, The salience net work was
73% larger on avera ge in the SIMD data set (signif icance asse ssed using a
permut ation test, *P = 0.001, B onferroni corre ction, Z-score = 6 .19). This effect
was replicated thrice (two-tailed independent sample t-test s, n = 48 from Weill
Cornell Me dicine, MDD -1: T = 3.54, *P = 0.01, Bonfe rroni correc tion, Cohen’s
d = 0.72; another s ample of n = 45 from Weill Cor nell Medicine , MDD-2: T = 4.17,
*P = 0.002, Bon ferroni correc tion, Cohen’s d = 0.8 4; n = 42 from Stanford
Universit y, MDD-3: T = 3.68, *P = 0.008, Bo nferroni correc tion, Cohen ’s d = 0.77).
Data are pre sented as mea n ± s.d. d, No signif icant group d ifferences in
salience n etwork repres entation in th e striatum were ob served in eith er the
discovery (two-tailed permutation test, P = 0.07, uncorrecte d) or replicatio n
datase ts (two-tailed in dependent s ample t-tests, all P > 0.43, un corrected).
Data are pre sented as mea n ± s.d. e, Densit y maps conf irm that spatial
locatio ns of salience ne twork nodes wer e similar in healthy c ontrols and
individual s with depres sion but that net work borders exte nded furth er
outwards f rom their centro ids in each cort ical zone in depre ssion (red boxes).
f, An SVM class ifier disti nguished indi viduals with de pression from h ealthy
controls ab ove chance (accuracy 78 .4%, signifi cance asse ssed using a
permut ation test, P = 0.001) u sing the size of eac h functional n etwork as
feature s. g, Linear predic tor coeff icients (β) ass ociated with t he trained mod el.
h, Change in mode l accuracy aft er exclusion of each n etwork. Both g a nd h
indicate th at salience ne twork size was the m ost import ant feature. ACC,
anterior ci ngulate; AI, an terior insular co rtex; Cd, cauda te; HC, healthy
controls; LPFC , lateral prefron tal; NAc, nucleus a ccumbens; PU, put amen;
SAL, salie nce network; S VM, support ve ctor machine .
Content courtesy of Springer Nature, terms of use apply. Rights reserved

Nature | Vol 633 | 19 September 2024 | 627
machine classifier to differentiate individuals with depression from
healthy control individuals on the basis of the size of all 20 functional
networks, pooling data from n = 37 healthy controls acquired from five
different scanners and n = 141 individuals with depression acquired from
two different scanners from two different manufacturers (that is, all
the data in Fig.1c). The Methods and Supplementary Fig.6 give details
about classifier training. Overall, support vector machine classifiers
correctly differentiated depression cases from healthy controls with
78.4% accuracy (permutation test, P = 0.001; Fig.1f), correctly identify-
ing 82.5% of depression cases, for a positive predictive value of 89.5%.
Feature importance was evaluated by examining the linear predictor
coefficients and calculating the change in accuracy after exclusion
(Fig.1g,h). As expected, salience network size was the most distinguish-
ing feature. Together, these analyses indicate that the salience network
is markedly expanded in most individuals with depression, with large
effect sizes that are reproducible in several samples involving different
data acquisition and analysis procedures, and sufficient in magnitude
to support individual classifications with high accuracy rates.
Three salience network expansion modes
Individual differences in functional brain organization occur in two
forms: ectopic intrusions, in which isolated pieces of a functional net-
work are observed in an atypical location, and border shifts, in which
the boundary of a network expands (or contracts) and encroaches on
its neighbours
15
. Border shifts are heritable
49
and associated with known
mechanisms of cortical expansion controlled by genetic programs that
refine boundaries between functional areas during development and
with experience or in response to environmental influences
50
. Further-
more, macroscale networks in both humans and non-human primates
are organized in a hierarchy associated with cortical gradients in gene
expression and functional properties
51
, with unimodal sensorimotor
areas at the base and heteromodal association areas such as the default
mode network at the apex
52
. Thus, as a step towards understanding the
mechanisms that give rise to cortical expansion of the salience network
in depression, we tested whether it was driven primarily by border
shifts or ectopic intrusions and whether it tended to affect lower-level,
unimodal sensorimotor networks or heteromodal association areas
positioned higher in this hierarchy. To this end, we first generated a
central tendency functional network map for the 37 healthy controls
(Fig.2a). Second, we identified parts of the salience network in each of
the 141 individuals with depression that did not overlap with the salience
network in the group-average map for healthy controls and classified
them as either ectopic intrusions or border shifts (Fig.2b,c). Next, we
calculated an encroachment profile for each subject, by quantifying the
degree of encroachment on every other functional network, defined
as the relative contribution of each functional network to the total sur-
face area of the encroaching portion of the salience network (Fig.2c).
This analysis confirmed that salience network expansion was not ran-
domly distributed—instead, it was due primarily to border shifts affecting
three neighbouring higher-order functional systems, with three distinct
encroachment profiles occurring in different individuals. Although sali-
ence network expansion involved both ectopic intrusions and border
shifts, the latter were more common (Fig.2d) and both tended to result
in encroachment on the default, frontoparietal or cingulo-opercular
networks (Fig.2e), not unimodal sensorimotor networks. Comparison
to 73 independent molecular, microstructural, electrophysiological,
developmental and functional brain maps from neuromaps toolbox
53
showed that salience network expansion frequently occurred in brain
regions with less intracortical myelin and thus greater capacity for
synaptic plasticity
54
and for which individual differences in functional
connectivity55 and the concentration of particular neurotransmitter
receptors (μ-opioid56 and histamine H3 receptors57) are most pronounced
(Extended Data Fig.6). Further comparisons to maps of FC test–retest
reliability and temporal signal-to-noise confirmed that these brain
regions were not more susceptible to noise than chance (Supplementary
Fig.7). It was also evident that the salience network tended to encroach
on specific functional networks in different cortical zones (Fig.2f). For
example, in the lateral prefrontal cortex, the salience network expanded
rostrally and tended to displace the frontoparietal network. By contrast,
in the anterior cingulate and anterior insular cortex, the default mode
and cingulo-opercular networks were disproportionately affected,
respectively. Clustering individuals by their encroachment profiles
showed three distinct modes (Fig.2g), involving predominantly the
default mode network, the frontoparietal network or a combination of
the frontoparietal and cingulo-opercular networks. This heterogeneity
may partly explain our observation that the salience network was consist-
ently expanded in all three datasets but corresponding contractions of
other functional networks were more variable.
The results above indicate that salience network expansion is driven
primarily by encroachment on the frontoparietal, cingulo-opercular
and default mode networks and suggest that cortical space at the
boundary between networks may be allocated to different functional
systems in individuals with depression. To test this, and to further vali-
date our findings, we compared the strength of functional connectivity
between encroaching nodes of the salience network (dark grey vertices
in the left part of Fig.2c) and the functional networks that typically
occupy that space in healthy controls. This analysis was performed
using split halves of each individual’s resting-state fMRI dataset to
assess the stability of the salience network assignment associated with
the encroaching vertices relative to the runner-up assignments (most
often either the default mode, frontoparietal or cingulo-opercular net-
work). As expected, the functional connectivity of encroaching salience
network nodes with the rest of the salience network was significantly
stronger (mean Z(r) = 0.26) than with the displaced networks (all mean
Z(r) < 0.12), consistent with weakened connectivity between encroach-
ing nodes and the functional networks that typically occupy that space
in healthy controls (Extended Data Fig.7). Together, these results show
that frontostriatal salience network expansion is driven primarily by
network border shifts that affect three specific higher-order functional
systems and spare others, with distinct modes of encroachment occur-
ring in three subgroups of patients.
Salience network topology is trait-like
Major depressive disorder is a fundamentally episodic condition
defined by discrete periods of low mood interposed between periods
of euthymia
58,59
. We evaluated if changes in salience network topology
accompany changes in the overall severity of depression symptoms that
occur during mood-state transitions—an hypothesis that our longitudi-
nal SIMD dataset was well-suited to test. However, and consistent with
previous work describing functional network topography in healthy
adults as very stable features affected very little by cognitive state
or daily variation
17,20
, we found that salience network topology was
stable over time in individuals with and without depression (Fig.3a).
Furthermore, within-subject analyses showed no significant correla
-
tion between fluctuations in depression symptoms (HDRS6, a more
sensitive measure of changes on shorter timescales) and changes in
salience network size over time in any of the densely sampled indi-
viduals in our SIMD dataset (Fig.3b). To address the same question,
we asked whether salience network size changed after a rapid acting
antidepressant treatment, leveraging samples of patients scanned
before and after a conventional 6-week course of repetitive transcranial
magnetic stimulation (rTMS; n = 90) or an accelerated, 1-week intensive
course of rTMS (n = 45). There was no significant pre-to-post change in
salience network size in either sample (Fig.3d). In addition, neither the
severity of symptoms during the current episode (Fig.3e) nor the total
number of depressive episodes individuals reported experiencing dur-
ing their lifetime (Fig.3f) explained individual differences in salience
network size. Collectively, these findings indicate that salience network
Content courtesy of Springer Nature, terms of use apply. Rights reserved

628 | Nature | Vol 633 | 19 September 2024
Article
expansion is a stable feature of individuals with major depressive dis-
order but not a marker of depressive episodes and unrelated to the
severity of their symptom severity or to the chronicity of their illness.
These observations led us to propose that instead of driving changes
in depressive symptoms over time, salience network expansion may be
a stable marker of risk for developing depression. To test this hypoth-
esis, we asked whether salience network expansion was present earlier
in life, before the onset of depressive symptoms in individuals. Using
data from the adolescent brain cognitive development (ABCD) study
60
,
we identified n = 57 children who did not have significant depressive
symptoms when they were scanned at ages 10 and 12 years but then went
on to develop clinically significant depressive symptoms at either age 13
or 14 years (Fig.3g). An equal number of children from the ABCD study
with no depressive symptoms at any time point were also identified as
a control sample. Precision functional mapping showed that, on average,
the salience network occupied 35.93% more of cortex in children with no
current or previous symptoms of depression at the time of their fMRI
scans but who subsequently developed clinically significant symptoms
of depression, relative to children with no depressive symptoms at any
study time point (Fig.3g, 3.81% ± 1.58% of cortex in ABCD-MDD versus
2.80% ± 1.48% of cortex in ABCD-HC). There was no significant change
in salience network size in the 2 years between the baseline and 2-year
follow-up visits in either sample (Supplementary Fig.8). A similar effect
was observed in adults with late-onset depression (Extended Data Fig.8).
Together, these results show that cortical expansion of the salience
network is a trait-like feature of brain network organization that is stable
over weeks, months and years, unaffected by mood state and detectable
in children before the onset of depression symptoms in adolescence.
Non-encroaching and encroaching parts of SAL (SIMD-4)HC group-average functional networks SAL in example patient (SIMD-4)
Striatum
ab
Cortical surface
DMN
SAL
CO
FP
c
Encroaching
(intrusions)CO
DMN
FP
Encroaching
(border shift)
Non-encroaching
eNetworks most
often displaced by SAL Networks displaced by SAL in different cortical zones
Anterior insula Medial prefrontal Lateral prefrontal
f
Encroachment (%)
50
Mode 1
0
g
DMN FP CO DMN FP CO DMN FP CO
SAL expansion affects different networks across individuals
Mode 2
CO
FP
DMN
Mode 3
0
25
50
75
100
Encroachment (%)
d
Expansion driven
by border shifts
*
0
20
40
60
80
Encroachment (%)
DMN FP CO
Border
shift
Ectopic
intrusion
0
15
30
45
60
Encroachment (%)
*
*
**
0
15
30
45
60
**
**
Mode 1Mode 2Mode 3
DMN FP CO DMN FP CO DMN FP CO
Encroachment (%)
Fig. 2 | Thre e modes of sa lience net work expansion i n depressi on. a, Mode
functio nal brain netwo rk assignmen ts in cortex and s triatum in HC . b,Salience
network in a re presentat ive individual (SIM D-4)with depress ion.c, The parts
of thesalienc e network of each i ndividual wit h depression th at did and did
not overlap with t he HC are referred to a s non-encroac hing and encroach ing,
respectively. d, Salience networ k expansion more of ten due to shift s in
network bo rders than ect opic intrusion s—isolated patch es of salience n etwork
in atypical locations (two-tailed paired sample t-test, *P < 0.001, n = 141). Data
are present ed as mean ± s.d . e, The parieta l subnetwork of th e DMN (red),
FP (yellow) and CO (pur ple) networks wer e most frequen tly displaced by
salience n etwork expansi on. f, Salience ne twork expansion a ffected dif ferent
functio nal networks i n different cor tical zones . In the AI, the FP (T = 5.94,
*P < 0.001) and CO (T = 6.42, *P < 0.001), netw orks were more affe cted than the
default mod e network. In th e ACC, the DMN was more af fected tha n either the
FP (T = 17.53, *P < 0.001) or CO/action-mo de (T = 15.25, *P < 0.0 01) networks.
Finally, in LPFC, the FP wa s more affecte d than either the D MN (T = 9.31,
*P < 0.001) or CO (T = 6.33, *P < 0.0 01). Statistical s ignifica nce was asses sed
using two-tailed two-sample t-test s; all P values are Bonferroni correct ed,
n = 141. Data are pres ented as mea n ± s.d. g, Individu als with depre ssion
clustered u sing their encro achment prof iles (the relative co ntribution of e ach
functio nal network to th e total surface a rea of the encroac hing portio n of their
salience n etwork) revealed t hree distinc t modes of encro achment acros s
individuals. CO, cingulo-opercular; DMN, default mo de; FP, frontoparietal.
Content courtesy of Springer Nature, terms of use apply. Rights reserved

Nature | Vol 633 | 19 September 2024 | 629
Connectivity state predicts anhedonia
The results above indicate that topological features of the salience
network, such as its size, shape and spatial location are stable over
time and do not fluctuate with mood state. However, this observa-
tion does not preclude the possibility that functional connectivity
between specific salience network nodes fluctuate in strength and that
such fluctuations contribute to the emergence of depressive episodes
and their subsequent remission. To test this, we first asked whether
changes in functional connectivity strength between nodes of the sali-
ence network either co-occur with or predict fluctuations in symptom
severity over time in individuals, focusing initially on hedonic function,
a core feature of depression that is associated with frontostriatal cir-
cuits
61–66
and is aligned with the putative role of salience network
40,41,46
and accumbens–anterior cingulate circuits more specifically67–69, in
reward processing and goal-oriented effortful behaviour
61,70–72
. Our
analyses focused on two of the patients from the SIMD dataset (SIMD-4
and SIMD-6) who were repeatedly scanned and assessed by clinicians
longitudinally over 8–18 months, providing sufficient data for this
analysis. This afforded an opportunity to ask for the first time at the
level of single densely sampled individuals—whose data effectively
served as independent, well-powered n-of-1 experiments
14,73
—how vari-
ability in brain network functional connectivity relates to fluctuations
in specific symptom domains.
Study visit 1
(depressed)
SIMD-1
(13 study visits, 6 months)
Study visit 5
(depressed)
Study visit 10
(not actively depressed)
Study visit 13
(depressed)
b
SAL size is stable over time within highly sampled individuals
cSIMD-1
(13 study visits, 6 months)
SIMD-4
(62 study visits, 1.5 years)
HDRS6
eNo effect of Sx
severity on SAL size
No effect of
chronicity on SAL size
fgSAL expansion before depression Sx onset
012345678910
Study visits
0
2
4
6
8
10
a
Healthy controls
12345678910
Study visits
0
2
4
6
8
10
Percentage
of cortex
Depression
HDRS6HDRS6 1-week Tx 6-week Tx
0510 15
0
2
4
6
8
10
Percentage
of cortex
SIMD-6
(39 study visits, 8 months)
d
r = –0.06
P = NS
No change in
SAL size post-rTMS
0510 15
0
2
4
6
8
10
r = –0.09
P = NS
0510 15
0
2
4
6
8
10
r = –0.06
P = NS
Remission
cutoff
Pre
Post
Pre
Post
0
2
4
6
8
10
NS NS
0510 15
0
2
4
6
8
10 r = 0.04
P = NS
0
2
4
6
8
10
HDRS6
<5 5–9 ≥10
Episodes reported
10 11 12 13 14
Age (years)
40
60
80
100
CBCL-depression
fMRI
fMRI
ABCD
HC
ABCD
MDD
Sx
onset
ABCD-HC
ABCD-MD
D
0
2
4
6
8
10
*
Percentage
of cortex
Percentage
of cortex
Percentage
of cortex
Percentage
of cortex
Percentage
of cortex
Percentage
of cortex
Percentage
of cortex
Fig. 3 | Sali ence network exp ansion is st able over time and p resent be fore
symptom onset. a, Cortical r epresenta tion of the salien ce network was s table
in repeate dly scanned h ealthy controls ( left) and indivi duals with depre ssion
(right). The f irst ten study v isits for each ind ividual are shown for v isualizatio n
purposes. b, Salience netwo rk in a represent ative individua l with depress ion
that was sc anned longit udinally to sample d ifferent mood s tates. c, No
signif icant correla tion betwee n the severity of d epressive sym ptoms (HDRS6)
and salien ce network size in a ny repeatedly s ampled indivi dual with depre ssion
from the SIM D sample (Pears on correlation, a ll P > 0.63, two-tail t est). d, No
signif icant change in s alience netwo rk size after a cou rse of either a tra ditional
6-week (two-tailed paired sample t-test, T = 0.58, P = 0. 55, uncorrec ted, n = 90)
or accelerated 1-week (two-tailed paired sample t-test, T = 0.58, P = 0.5 6,
uncorrected, n = 45) co urse of repeti tive transcran ial magnetic st imulation
(rTMS). Data are plotted a s mean ± s.d. e, Indi vidual differen ces in salienc e
network size were not significantly correlated with depres sion severity
(HDRS6, Pe arson correla tion, r = 0.04, P = 0.63, uncor rected, two -tailed test).
f,The number of de pressive episo des experie nced (inferred f rom the
Mini-Inter national Neur opsychiatric In terview) in eac h individual’s lifeti me
in relation to t he size of theirsalien ce network. D ata are plotte d as mean ± s.d.
g, Children from th e ABCD study sc anned before the o nset of elevate d
depressi on symptoms were i dentifie d (ABCD-MD D). Depression sym ptoms
were operat ionalized usin g the DSM-orie nted scale for dep ression from th e
CBCL (T-scores ≥70 are in the clinical r ange). The salience n etwork was
significantly larger in children who later developed clinically elevated
symptoms o f depression co mpared to children w ho did not (two-taile d
independent sample t-test, T = 3.50, *P < 0.0 01, Cohen’s d = 0.62, n = 114). Data
are plotte d as mean ± s.d. NS , not signif icant; Sx, sympto m; Tx, treatment.
Content courtesy of Springer Nature, terms of use apply. Rights reserved

630 | Nature | Vol 633 | 19 September 2024
Article
We began with SIMD-4 because this individual was studied over the
longest period of time (62 study visits over 1.5 years) and had the most
fMRI data (29.96 h of fMRI data in total) and reserved SIMD-6 as a repli-
cation dataset (57 study visits over 12 months, the initial 39 study visits
had fMRI data before DBS implantation, 18.85 h of fMRI data in total).
During a period spanning 1.5 years, we observed significant fluctuations
in ten anhedonia-related measures (Fig.4a), which were derived from
five standardized depressive symptom scales and identified by a con-
sensus clinical decision by three study co-authors (Supplementary
Fig.9), ranging from mild or negligible to severe. We tested whether
changes in functional connectivity between nodes of the salience
network were correlated with changes in anhedonia in this individual
NAc LH
NAc RH
Cd LH
Cd RH
PU LH
PU RH
–1
0
1
PC1
8/18/20
8/25/20
9/01/20
9/08/20
9/15/20
9/22/20
9/29/20
10/06/20
10/13/20
10/20/20
10/27/20
11/03/20
11/10/20
11/17/20
11/24/20
12/01/20
12/08/20
12/15/20
12/22/20
12/29/20
1/05/21
1/12/21
1/19/21
1/26/21
2/02/21
2/09/21
2/16/21
2/23/21
3/02/21
3/09/21
3/16/21
3/23/21
3/30/21
4/06/21
4/13/21
4/20/21
4/27/21
5/04/21
5/11/21
5/18/21
5/25/21
6/01/21
6/08/21
6/15/21
6/22/21
6/29/21
7/06/21
7/13/21
7/20/21
7/27/21
8/03/21
8/10/21
8/17/21
8/24/21
8/31/21
9/07/21
9/14/21
9/21/21
9/28/21
10/05/21
10/12/21
10/19/21
10/26/21
11/02/21
11/09/21
11/16/21
11/23/21
11/30/21
12/07/21
12/14/21
HDRS-Activities
HDRS-Libido
STAI-Satised
PANAS-Interested
PANAS-Enthusiastic
QIDS-Interest
BDI-Anhedonia
BDI-Interest
BDI-Libido
PSQI-Enthusiasm
Corticostriatal SAL
nodes in SIMD-4
b
PU
NAc
Cd
x = 10
LPFC
ACC
AI
SIMD-4SIMD-6
c
Time
0.40
r
NS
Corticostriatal SAL FC
correlations with anhedonia
d
Anxiety (NAc <-> AI SAL FC)
Anxiety (PC1) Anxiety (PC1)
aLong-term (62 study visits over 1.5 years) assessment of anhedonia in SIMD-4
Sx severity
Maximum
Minimum
–0.40
Prediction of future anhedonia Sx
(NAc <-> ACC SAL FC)
e
h
SHAPS total score
fAll depression
(n = 135, cross-sectional)
20 30 40 50
0
0.1
0.2
0.3
0.4
0.5
NAc <--> ACC SAL FC
r = 0.09
P = NS
–2 –1 012
0
0.1
0.2
0.3
0.4
0.5
NAc <--> ACC SAL FC
Anhedonia (NAc <-> ACC SAL FC)
–2 –1 012
–0.50
–0.25
0
0.25
0.50
r
Null
distribution
**
SIMD-4
–2 –1 012
0
0.1
0.2
0.3
0.4
0.5
NAc <--> ACC SAL FC
SIMD-4 r = –0.16
P = NS
Anxiety (PC1)
Lag
–0.50
–0.25
0
0.25
0.50
r
SIMD-6
FC predicting
past symptoms
FC predicting
future symptoms (study visits)
–2 –1 012
0
0.15
0.30
0.45
0.60
0.75
Anxiety (PC1)
–2 –1 012
0
0.15
0.30
0.45
0.60
0.75
NAc <--> AI SAL FC
SIMD-4
r = –0.29
P = 0.02
SIMD-6
r = –0.45
P = 0.003
–2 –1 012
0
0.1
0.2
0.3
0.4
0.5
r = –0.28
P = NS
SIMD-6
–2 –1 012
0
0.1
0.2
0.3
0.4
0.5
SIMD-4
r = –0.37
P = 0.003
SIMD-6
r = –0.49
P = 0.001
ACC LH
ACC RH
LPFC LH
LPFC RH
AI LH
AI RH
NAc LH
NAc RH
Cd LH
Cd RH
PU LH
PU RH
Anhedonia (PC1)
gAnxiety (NAc <-> ACC SAL FC)
Anhedonia (PC1)
*
–2 –1 012
Fig. 4 | Frontos triatal sa lience net work connect ivity pred icts f luctuatio ns
in anhedonia and anxiety symptoms in deeply sampled individuals with
depression over time. a, Heat map summariz ing fluc tuations in indi vidual
items sel ected from a var iety of clinic al interview s and self-report sc ales
related to anh edonia sympto ms in a deeply sam pled individu al with depres sion
(SIMD-4). Clinic al data were resam pled to days for visual ization purp oses
(black dot s mark the study vi sits). b, Frontostriatal n odes of the sali ence
network in SI MD-4. c, Corre lation matric es summariz ing the associ ation
betwee n FC strength be tween differ ent cortico -striatal s alience netwo rk
nodes and f luctuat ions in the severi ty of anhedoni a-related sympto ms in
both SIMD -4 and SIMD -6. d, FC betwe en salience ne twork nodes in th e NAc
and ACC most clos ely tracked f luctuation s in the severit y of anhedonia-rela ted
symptoms i n both SIMD- 4 (Pearson corre lation, r = −0. 37, P = 0.003) and
SIMD- 6 (Pearson corre lation, r = −0.49, P = 0.001) across study v isits. Sta tistical
signif icance was as sessed usin g two-tailed p ermutatio n tests with c ircular
rotation to p reserve temp oral autocorre lation. e, Cross- correlation an alyses
indicate d NAc ←→ ACC FC also predic ted the severit y of anhedonia-re lated
symptoms a t the following stud y visit in SIMD -4 (Pearson cor relation,
signif icance test ed by means of per mutation te st, r = −0.32, *P = 0.0 04) but
not in SIMD -6. f, No signif icant corre lation betwe en individual di fferences in
salience n etwork NAc ←→ ACC FC stre ngth and the s everity of anhe donia-related
symptoms a cross indiv iduals (assess ed using SHA PS, Pears on correlatio n,
r = 0.09, P = 0.41). g, In SIMD- 4 and SIMD-6, s alience net work NAc ←→ ACC FC
was not sign ificantly re lated to flu ctuations in t he severity of ot her depressive
symptoms , such as anxiet y. h, In contrast, FC bet ween theNAc and A I was most
closely rel ated to fluc tuations in th e severity of anx iety-related s ymptoms
(Pearson co rrelations, SI MD-4: r = −0. 29, P = 0.02; SIMD-6: r = − 0.45, P = 0.004,
two-tail ed tests), indicati ng different pat terns of func tional conne ctivity
relate to dif ferent symptom s. FC, functi onal connect ivity. AI, ante rior insula;
NAc, nucleus accumbens.
Content courtesy of Springer Nature, terms of use apply. Rights reserved

Nature | Vol 633 | 19 September 2024 | 631
over time, as measured by a principal component analysis of the ten
anhedonia-related measures in Fig.4a and summarized by the first com-
ponent score. We found that functional connectivity between several
cortical and striatal salience network nodes was correlated with changes
in anhedonia over time (Fig.4c,d), with the strongest effects observed
for connectivity between the nucleus accumbens and anterior cingulate
cortex. An identical analysis in SIMD-6, involving 39 study visits with
clinical and fMRI data over 8 months, replicated this effect (Fig.4c,d
and Extended Data Fig.9a,b). This finding remained significant in both
individuals when including head motion at each study visit as a covariate.
Next, we asked whether salience network functional connectivity
was predictive of symptom severity at future study visits and whether
the effect was specific to anhedonia or extended to other symptom
domains. Notably, a cross-correlation analysis examining correlations
with symptoms in past, present and future study visits showed that
functional connectivity between the salience network nodes in the
nucleus accumbens and anterior cingulate was not only correlated
with current anhedonia symptoms but also predicted the future emer-
gence or remission of anhedonia symptoms in the next study visit in
SIMD-4 (Fig.4e, top), typically with a lag of approximately 1 week. The
significance of this effect was confirmed using permutation tests with
circular rotation to preserve temporal autocorrelation, indicating that
accumbens–anterior cingulate connectivity at a given visit predicted
future anhedonia approximately 1 week later, even after controlling
for correlations in anhedonia measures over time. Of note, salience
network connectivity correlations were replicated in SIMD-6 for cur-
rent symptoms but not for future symptoms (Fig.4e, bottom), which
may relate to differences in their antidepressant treatments in this
observational setting (SIMD-6 was undergoing maintenance electro-
convulsive treatment unrelated to this study).
To determine whether changes in nucleus accumbens–anterior cin-
gulate functional connectivity not only predicted changes in anhedonia
in individual subjects over time but also explained individual differ-
ences in anhedonia at a given point in time, we repeated this analysis
cross-sectionally using the entire n = 135 cohort of replication subjects
using a standardized self-report measure of anhedonia. However, this
analysis did not show a significant correlation between individual dif-
ferences in functional connectivity between the anterior cingulate
and nucleus accumbens and anhedonia across individuals (Fig.4f),
underscoring the value of within-subject analyses.
Finally, to evaluate the specificity of this effect, we asked whether
nucleus accumbens–anterior cingulate connectivity was also associ-
ated with anxiety, a symptom domain that co-occurs with depression
but is often dissociable from anhedonia (see Extended Data Fig.10 for
stacked anhedonia and anxiety symptom heatmaps). For example,
‘dysphoric’ (sadness and anhedonia) and ‘anxiosomatic’ (anxiety and
somatic) symptoms were dissociable from one another in a recent study
mapping response to rTMS intervention to different stimulation sites
74
.
We did not observe a significant correlation between accumbens–
anterior cingulate connectivity and anxiety in either individual (Fig.4g),
indicating a more important role for this circuit in anhedonia. Of
note, there are several neuroimaging and circuit physiology studies
implicating the insula in the expression of anxiety and the process-
ing of aversive states
75–79
. Motivated by this work, we performed an
analogous analysis asking whether changes in striatal connectivity
with the anterior insula area of the salience network were correlated
with fluctuations in anxiety symptoms over time in each subject. In
accord with our prediction, we found that striatal connectivity with
anterior insula was significantly correlated with anxiety symptoms in
SIMD-4 and replicated this effect in SIMD-6 (Fig.4h). An exploratory
whole-brain analysis evaluating how salience network connectivity
strength to the rest of the cortex relates to fluctuations in the severity of
anhedonia and anxiety symptoms is summarized in Supplementary
Fig.10. Collectively, these findings show that, although the salience
network is stably expanded in individuals with depression and that this
expansion seems to occur early in life, frontostriatal connectivity in this
network also fluctuates over time, and changes in striatal connectivity
with the anterior cingulate and anterior insula track the emergence and
remission of anhedonia and anxiety symptoms, respectively.
Interpreting differences in topology
In this work, precision functional mapping in deeply sampled individuals
with depression showed a marked expansion of the salience network that
was robust and reproducible in several samples, with medium to large
effect sizes relative to previously reported neuroimaging abnormalities
in depression. This effect was driven primarily by network border shifts
that encroached on three specific functional systems—the frontopari-
etal, cingulo-opercular and default mode networks—with three distinct
modes of encroachment in different individuals. This effect was stable
over time, not sensitive to mood state or a marker of depressive episodes
and emerged early in life in children who went on to develop depressive
symptoms later in adolescence. At the same time, changes in striatal
connectivity with anterior cingulate and anterior insula nodes of the
salience network tracked the emergence and remission of anhedonia
and anxiety, respectively, and predicted future changes in hedonic
function in one individual. Of note, our analysis benefited from the use
of precision functional mapping in combination with large quantities
of high-quality, densely sampled multi-echo fMRI data, which may be
critical for mapping individual differences in network topology precisely
(Supplementary Fig.12) and this might in part explain why these findings
have not been reported in the literature previously.
Although more work will be required to elucidate the mechanisms
underlying salience network expansion in depression, key results from
this report and other studies point to at least two hypotheses. First,
converging evidence from several sources indicates that individual
differences in network topology are regulated by activity-dependent
mechanisms and related to the extent to which a given network is
actively used. So far, most studies evaluating variability in the size of
functional areas or networks across individual humans or other animals
have focused primarily on the motor and visual systems. These studies
have shown how different body parts have distinct representations in
the primary motor cortex (M1) that differ in size and cortical repre-
sentation is closely related to the dexterity of the corresponding limb,
such that the upper limbs occupy more cortical surface area than the
lower limbs, as one example
80
. Motor training can increase the rep-
resentation of the trained muscle or limb in M1 (refs. 81,82), whereas
limb amputation, casting and congenital limb defects all decrease the
representation of the disused limb and increase the representation of
other body parts
83–85
. The total surface area of primary visual cortex (V1)
can vary up to threefold in healthy young adults and is correlated with
individual differences in visual awareness86 and contrast sensitivity87.
Likewise, total cortical representation of the frontoparietal network
was found to be positively correlated with executive function abilities
in children88. Together, these reports suggest that salience network
expansion—accompanied by a corresponding contraction of the fron-
toparietal, cingulo-opercular or default mode networks—may reflect a
reallocation of cortical territory and information processing priorities
in individuals with depression, which could in turn contribute to altera
-
tions in salience network functions such as interoceptive awareness,
reward learning, autonomic signal processing and effort valuation30,40,41.
Second, converging data indicate that cortical network topology
is strongly influenced not only by externally modulated, activity-
dependent mechanisms but also by intrinsic genetic programs
50,89
.
Numerous transcription factors regulate cell adhesion molecules,
exhibit strong expression gradients across the cortical sheet during
development and covary with aspects of cortical organization, includ-
ing the size or location of functional areas
89,90
. Deletion of these pat-
terning factors can result in contraction or expansion of functional
areas
90
. Conversely, increased expression of Emx2 increases the size
Content courtesy of Springer Nature, terms of use apply. Rights reserved

632 | Nature | Vol 633 | 19 September 2024
Article
of V1 and decreases the size of somatomotor areas91,92. Although our
findings do not speak directly to this question, at least three observa-
tions are consistent with a role for intrinsic developmental genetic
programs as opposed to exclusively activity-dependent mechanisms.
First, salience network expansion was highly stable, irrespective of an
individual’s current mood state, indicating that acute mood-related
changes in network activity did not influence network size. Second,
salience network expansion emerged early in life, consistent with a
developmentally regulated mechanism. And third, salience network
expansion was driven by spatially organized border shifts, which are
known to be heritable
19,93
and which tended to encroach on neighbour-
ing networks in specific directions, expanding anteriorly and dispro-
portionately targeting higher-order heteromodal association cortex
although sparing unimodal sensorimotor areas.
Trait versus state effects in depression
Our findings may also open new avenues to addressing two fundamen
-
tal challenges to using insights from clinical neuroimaging research
to rethink our approach to diagnosing and treating depression. First,
as noted above, MRI studies spanning two decades have identified
anatomical and functional connectivity alterations that are robust and
reproducible in large-scale meta-analyses but are highly variable across
subjects with modest effect sizes (typically Cohen’s d = 0.10–0.35),
which complicates effects to leverage these effects for clinical pur-
poses. By contrast, salience network expansion was observed in most
individuals with depression in our sample, readily apparent on visual
inspection and associated with medium to large effect sizes (Cohen’s d =
0.77–1.99). This effect was detected without corrections for site- or
scanner-induced biases—which can be a significant confound for mul-
tisite neuroimaging data.
Biomarkers in several areas of medicine come in different forms,
some of which are sensitive to current symptoms, whereas others are
stable trait-like markers of disease or a marker of risk for developing
symptoms. Our study was not designed to comprehensively validate a
neuroimaging biomarker for depression and future work will be needed
to assess the specificity of our findings with respect to other forms of
psychopathology or evaluate its potential clinical utility. A preliminary
analysis indicated that the salience network is also larger than normal
in two individuals with bipolar II disorder but not autism spectrum
disorder or obsessive compulsive disorder (Supplementary Fig.11),
which might reflect common deficits in behavioural domains, such as
reward processing
94
, that are also linked to salience network function.
However, our results do indicate that salience network expansion has
the potential to help predict susceptibility to depression symptoms
and could have important implications for designing therapeutic neu-
romodulation interventions, which could have widely varying effects
due to individual differences in network topology
95,96
. Another caveat
to keep in mind is that precise and reliable mapping of the salience
network and consistent detection of salience network expansion in
individuals with depression may require 1.5–2 h of high-quality fMRI
data per subject (Supplementary Fig.12), which may be an obstacle for
retrospective analysis of traditional fMRI datasets not optimized for
precision functional mapping at the individual level. It is also notewor-
thy that the brain network that we and others3,41,45,46 have referred to
as the salience network is sometimes called other names (Control C in
ref. 97) or combined with the parietal memory network
98
, whereas the
brain network we refer to as cingulo-opercular/action-mode network
41
is sometimes called the salience/ventral attention network
97
(Supple-
mentary Fig. 13). Developing a standardized functional brain network
nomenclature
99
will improve the interpretability of insights gleaned
from precision functional mapping such as in the present study.
Finally, our study provides proof-of-principle data to support the
use of precision functional mapping and deep, longitudinal sampling
for understanding cause and effect in clinical neuroimaging studies of
depression. Our analyses show stable, trait-like differences in salience
network topology that are not only associated with depression but also
emerge early in life in children with no history of depression and predict
the subsequent emergence of depressive symptoms in adolescence.
At the same time, they show how changes in functional connectivity
strength between specific salience network nodes track the emer-
gence and remission of dysfunction in specific symptom domains in
individuals over time and, in at least one individual, predict the future
emergence of anhedonia symptoms at least 1 week before they occur. In
this way, they show how dense sampling and longitudinal designs will
open new avenues for understanding cause and effect and for design-
ing personalized, prophylactic treatments.
Online content
Any methods, additional references, Nature Portfolio reporting summa-
ries, source data, extended data, supplementary information, acknowl-
edgements, peer review information; details of author contributions
and competing interests; and statements of data and code availability
are available at https://doi.org/10.1038/s41586-024-07805-2.
1. Winter, N. R. etal. Quantifying deviations of brain structure and function in major
depressive disorder across neuroimaging modalities. JAMA Psychiatry 79, 879–888 (2022).
2. Laumann, T. O. etal. Functional system and areal organization of a highly sampled
individual human brain. Neuron 87, 657–670 (2015).
3. Gordon, E. M. etal. Precision functional mapping of individual human brains. Neuron 95,
791–807 (2017).
4. Braga, R. M. & Buckner, R. L. Parallel interdigitated distributed networks within the
individual estimated by intrinsic functional connectivity. Neuron 95, 457–471 (2017).
5. Duman, R. S. & Aghajanian, G. K. Synaptic dysfunction in depression: potential
therapeutic targets. Science 338, 68–72 (2012).
6. Duman, R. S., Aghajanian, G. K., Sanacora, G. & Krystal, J. H. Synaptic plasticity and
depression: new insights from stress and rapid-acting antidepressants. Nat. Med. 22,
238–249 (2016).
7. Lui, S. etal. Resting-state functional connectivity in treatment-resistant depression.
Am. J. Psychiatry 168, 642–648 (2011).
8. Williams, L. M. Precision psychiatry: a neural circuit taxonomy for depression and anxiety.
Lancet Psychiatry 3, 472–480 (2016).
9. Drysdale, A. T. etal. Resting-state connectivity biomarkers deine neurophysiological
subtypes of depression. Nat. Med. 23, 28–38 (2017).
10. Murray, C. J. L. etal. Disability-adjusted life years (DALYs) for 291 diseases and injuries in 21
regions, 1990–2010: a systematic analysis for the Global Burden of Disease Study 2010.
Lancet 380, 2197–2223 (2012).
11. Wang, D. etal. Parcellating cortical functional networks in individuals. Nat. Neurosci. 18,
1853–1860 (2015).
12. Poldrack, R. A. etal. Long-term neural and physiological phenotyping of a single human.
Nat. Commun. 6, 8885 (2015).
13. Fedorenko, E. The early origins and the growing popularity of the individual-subject
analytic approach in human neuroscience. Curr. Opin. Behav. Sci. 40, 105–112 (2021).
14. Naselaris, T., Allen, E. & Kay, K. Extensive sampling for complete models of individual
brains. Curr. Opin. Behav. Sci. 40, 45–51 (2021).
15. Seitzman, B. A. etal. Trait-like variants in human functional brain networks. Proc. Natl
Acad. Sci. USA 116, 22851–22861 (2019).
16. Kraus, B. T. etal. Network variants are similar between task and rest states. Neuroimage
229, 117743 (2021).
17. Gratton, C. etal. Functional brain networks are dominated by stable group and individual
factors, not cognitive or daily variation. Neuron 98, 439–452 (2018).
18. Lynch, C. J. etal. Rapid precision functional mapping of individuals using Multi-Echo fMRI.
Cell Rep. 33, 108540 (2020).
19. Dworetsky, A. etal. Two common and distinct forms of variation in human functional brain
networks. Nat. Neurosci. 27, 1187–1198 (2024).
20. Anderson, K. M. etal. Heritability of individualized cortical network topography. Proc. Natl
Acad. Sci. USA 118, e2016271118 (2021).
21. Parsons, S. & McCormick, E. M. Limitations of two time point data for understanding
individual differences in longitudinal modeling—what can difference reveal about
change? Dev. Cogn. Neurosci. 66, 101353 (2024).
22. Kraus, B. etal. Insights from personalized models of brain and behavior for identifying
biomarkers in psychiatry. Neurosci. Biobehav. Rev. 152, 105259 (2023).
23. Sendi, M. S. E. etal. Intraoperative neural signals predict rapid antidepressant effects of
deep brain stimulation. Transl. Psychiatry 11, 551 (2021).
24. Laumann, T. O. etal. Brain network reorganisation in an adolescent after bilateral perinatal
strokes. Lancet Neurol. 20, 255–256 (2021).
25. Dohm, K., Redlich, R., Zwitserlood, P. & Dannlowski, U. Trajectories of major depression
disorders: a systematic review of longitudinal neuroimaging indings. Aust. NZ J.
Psychiatry 51, 441–454 (2017).
26. Brady, R. O. et al. Bipolar mood state relected in cortico-amygdala resting state
connectivity: a cohort and longitudinal study. J. Affect. Disord. 217, 205–209 (2017).
27. Rey, G. et al. Dynamics of amygdala connectivity in bipolar disorders: a longitudinal study
across mood states. Neuropsychopharmacology 46, 1693–1701 (2021).
Content courtesy of Springer Nature, terms of use apply. Rights reserved

Nature | Vol 633 | 19 September 2024 | 633
28. Scangos, K. W. etal. Closed-loop neuromodulation in an individual with treatment-resistant
depression. Nat. Med. 27, 1696–1700 (2021).
29. Alagapan, S. etal. Cingulate dynamics track depression recovery with deep brain
stimulation. Nature 622, 130–138 (2023).
30. Seeley, W. W. etal. Dissociable intrinsic connectivity networks for salience processing
and executive control. J. Neurosci. 27, 2349–2356 (2007).
31. Kundu, P. etal. Multi-echo fMRI: a review of applications in fMRI denoising and analysis of
BOLD signals. Neuroimage 154, 59–80 (2017).
32. Mayberg, H. S. etal. Reciprocal limbic-cortical function and negative mood: converging
PET indings in depression and normal sadness. Am. J. Psychiatry 156, 675–682 (1999).
33. Ressler, K. J. & Mayberg, H. S. Targeting abnormal neural circuits in mood and anxiety
disorders: from the laboratory to the clinic. Nat. Neurosci. 10, 1116–1124 (2007).
34. Goodkind, M. etal. Identiication of a common neurobiological substrate for mental
illness. JAMA Psychiatry 72, 305–315 (2015).
35. Kaiser, R. H., Andrews-Hanna, J. R., Wager, T. D. & Pizzagalli, D. A. Large-scale network
dysfunction in major depressive disorder: a meta-analysis of resting-state functional
connectivity. JAMA Psychiatry 72, 603–611 (2015).
36. Schmaal, L. etal. Cortical abnormalities in adults and adolescents with major depression
based on brain scans from 20 cohorts worldwide in the ENIGMA Major Depressive
Disorder Working Group. Mol. Psychiatry 22, 900–909 (2017).
37. Mayberg, H. S. etal. Deep brain stimulation for treatment-resistant depression. Neuron
45, 651–660 (2005).
38. Crowell, A. L. etal. Long-term outcomes of subcallosal cingulate deep brain stimulation
for treatment-resistant depression. Am. J. Psychiatry 176, 949–956 (2019).
39. Hamilton, M. A rating scale for depression. J. Neurol. Neurosurg. Psychiatry 23, 56–62 (1960).
40. Seeley, W. W. The salience network: a neural system for perceiving and responding to
homeostatic demands. J. Neurosci. 39, 9878–9882 (2019).
41. Dosenbach, N. U. F., Raichle, M. E. & Gordon, E. M. The brain’s cingulo-opercular
action-mode network. Preprint at PsyArXiv https://doi.org/10.31234/osf.io/2vt79 (2024).
42. Kong, R. etal. Spatial topography of individual-speciic cortical networks predicts human
cognition, personality, and emotion. Cereb. Cortex 29, 2533–2551 (2019).
43. Haber, S. N. & Knutson, B. The reward circuit: linking primate anatomy and human
imaging. Neuropsychopharmacology 35, 4–26 (2010).
44. Shepherd, G. M. G. Corticostriatal connectivity and its role in disease. Nat. Rev. Neurosci.
14, 278–291 (2013).
45. Power, J. D. etal. Functional network organization of the human brain. Neuron 72,
665–678 (2011).
46. Gordon, E. M. etal. Individualized functional subnetworks connect human striatum and
frontal cortex. Cereb. Cortex 32, 2868–2884 (2022).
47. Van Essen, D. C. etal. The WU-Minn Human Connectome Project: an overview.
Neuroimage 80, 62–79 (2013).
48. Blumberger, D. M. etal. Effectiveness of theta burst versus high-frequency repetitive
transcranial magnetic stimulation in patients with depression (THREE-D): a randomised
non-inferiority trial. Lancet 391, 1683–1692 (2018).
49. Dworetsky, A. etal. Two common and distinct forms of variation in human functional
brain networks. Nat. Neurosci. https://doi.org/10.1038/s41593-024-01618-2 (2024).
50. Krubitzer, L. A. & Seelke, A. M. H. Cortical evolution in mammals: the bane and beauty of
phenotypic variability. Proc. Natl Acad. Sci. USA 109, 10647–10654 (2012).
51. Wang, X.-J. Macroscopic gradients of synaptic excitation and inhibition in the neocortex.
Nat. Rev. Neurosci. 21, 169–178 (2020).
52. Margulies, D. S. etal. Situating the default-mode network along a principal gradient of
macroscale cortical organization. Proc. Natl Acad. Sci. USA 113, 12574–12579 (2016).
53. Markello, R. D. etal. neuromaps: structural and functional interpretation of brain maps.
Nat. Methods 19, 1472–1479 (2022).
54. Glasser, M. F., Goyal, M. S., Preuss, T. M., Raichle, M. E. & Van Essen, D. C. Trends and
properties of human cerebral cortex: correlations with cortical myelin content.
Neuroimage 93, 165–175 (2014).
55. Mueller, S. etal. Individual variability in functional connectivity architecture of the human
brain. Neuron 77, 586–595 (2013).
56. Turtonen, O. etal. Adult attachment system links with brain mu opioid receptor
availability invivo. Biol. Psychiat. Cogn. Neurosci. Neuroimaging 6, 360–369 (2021).
57. Gallezot, J.-D. etal. Kinetic modeling of the serotonin 5-HT(1B) receptor radioligand [(11)C]
P943 in humans. J. Cereb. Blood Flow Metab. 30, 196–210 (2010).
58. Post, R. M. etal. Morbidity in 258 bipolar outpatients followed for 1 year with daily
prospective ratings on the NIMH life chart method. J. Clin. Psychiatry 64, 680–690 (2003).
59. Malhi, G. S. & Mann, J. Depression. Lancet 392, 2299–2312 https://doi.org/10.1016/s0140-
6736(18)31948-2 (2018).
60. Casey, B. J. etal. The adolescent brain cognitive development (ABCD) study: imaging
acquisition across 21 sites. Dev. Cogn. Neurosci. 32, 43–54 (2018).
61. Kennerley, S. W., Walton, M. E., Behrens, T. E. J., Buckley, M. J. & Rushworth, M. F. S. Optimal
decision making and the anterior cingulate cortex. Nat. Neurosci. 9, 940–947 (2006).
62. Rushworth, M. F. S. & Behrens, T. E . J. Choice, uncertainty and value in prefrontal and
cingulate cortex. Nat. Neurosci. 11, 389–397 (2008).
63. Lim, B. K., Huang, K. W., Grueter, B. A., Rothwell, P. E. & Malenka, R. C. Anhedonia requires
MC4R-mediated synaptic adaptations in nucleus accumbens. Nature 487, 183–189 (2012).
64. Russo, S. J. & Nestler, E. J. The brain reward circuitry in mood disorders. Nat. Rev.
Neurosci. 14, 609–625 (2013).
65. Tye, K. M. etal. Dopamine neurons modulate neural encoding and expression of
depression-related behaviour. Nature 493, 537–541 (2013).
66. Pizzagalli, D. A. Depression, stress, and anhedonia: toward a synthesis and integrated
model. Annu. Rev. Clin. Psychol. 10, 393–423 (2014).
67. Lee, D., Seo, H. & Jung, M. W. Neural basis of reinforcement learning and decision making.
Annu. Rev. Neurosci. 35, 287–308 (2012).
68. Ruff, C. C. & Fehr, E. The neurobiology of rewards and values in social decision making.
Nat. Rev. Neurosci. 15, 549–562 (2014).
69. Fetcho, R. N. etal. A stress-sensitive frontostriatal circuit supporting effortful reward-
seeking behavior. Neuron 112, 473–487 (2024).
70. Hart, E. E., Stolyarova, A., Conoscenti, M. A., Minor, T. R. & Izquierdo, A. Rigid patterns of
effortful choice behavior after acute stress in rats. Stress 20, 19–28 (2017).
71. Quilodran, R., Rothé, M. & Procyk, E. Behavioral shifts and action valuation in the anterior
cingulate cortex. Neuron 57, 314–325 (2008).
72. Prévost, C., Pessiglione, M., Météreau, E., Cléry-Melin, M.-L. & Dreher, J.-C. Separate
valuation subsystems for delay and effort decision costs. J. Neurosci. 30, 14080–14090
(2010).
73. Newbold, D. J. & Dosenbach, N. U. F. Tracking plasticity of individual human brains.
Curr. Opin. Behav. Sci. 40, 161–168 (2021).
74. Siddiqi, S. H. etal. Distinct symptom-speciic treatment targets for circuit-based
neuromodulation. Brain Stim. 12, e138 https://doi.org/10.1016/j.brs.2019.03.052
(2019).
75. Terasawa, Y., Shibata, M., Moriguchi, Y. & Umeda, S. Anterior insular cortex mediates
bodily sensibility and social anxiety. Soc. Cogn. Affect. Neurosci. 8, 259–266 (2013).
76. Gogolla, N. The insular cortex. Curr. Biol. 27, R580–R586 (2017).
77. Deng, H. etal. A genetically deined insula-brainstem circuit selectively controls
motivational vigor. Cell 184, 6344–6360 (2021).
78. Nicolas, C. etal. Linking emotional valence and anxiety in a mouse insula–amygdala
circuit. Nat. Commun. 14, 5073 (2023).
79. Hsueh, B. etal. Cardiogenic control of affective behavioural state. Nature 615, 292–299
(2023).
80. Penield, W. & Boldrey, E. Somatic motor and sensory representation in the cerebral
cortex of man as studied by electrical stimulation. Brain 60, 389–443 (1937).
81. Karni, A. etal. Functional MRI evidence for adult motor cortex plasticity during motor skill
learning. Nature 377, 155–158 (1995).
82. Taubert, M., Mehnert, J., Pleger, B. & Villringer, A. Rapid and speciic gray matter changes
in M1 induced by balance training. Neuroimage 133, 399–407 (2016).
83. Yu, X. J. etal. Somatotopic reorganization of hand representation in bilateral arm
amputees with or without special foot movement skill. Brain Res. 1546, 9–17 (2014).
84. Hahamy, A. etal. Representation of multiple body parts in the missing-hand territory of
congenital one-handers. Curr. Biol. 27, 1350–1355 (2017).
85. Nakagawa, K., Takemi, M., Nakanishi, T., Sasaki, A. & Nakazawa, K. Cortical reorganization
of lower-limb motor representations in an elite archery athlete with congenital
amputation of both arms. Neuroimage Clin. 25, 102144 (2020).
86. Schwarzkopf, D. S., Song, C. & Rees, G. The surface area of human V1 predicts the
subjective experience of object size. Nat. Neurosci. 14, 28–30 (2011).
87. Himmelberg, M. M., Winawer, J. & Carrasco, M. Linking individual differences in human
primary visual cortex to contrast sensitivity around the visual ield. Nat. Commun. 13,
3309 (2022).
88. Cui, Z. etal. Individual variation in functional topography of association networks in
youth. Neuron 106, 340–353 (2020).
89. Cadwell, C. R., Bhaduri, A., Mostajo-Radji, M. A., Keefe, M. G. & Nowakowski, T. J.
Development and arealization of the cerebral cortex. Neuron 103, 980–1004 (2019).
90. O’Leary, D. D. & Sahara, S. Genetic regulation of arealization of the neocortex. Curr. Opin.
Neurobiol. 18, 90–100 (2008).
91. Hamasaki, T., Leingärtner, A., Ringstedt, T. & O’Leary, D. D. M. EMX2 regulates sizes and
positioning of the primary sensory and motor areas in neocortex by direct speciication of
cortical progenitors. Neuron 43, 359–372 (2004).
92. Leingärtner, A. etal. Cortical area size dictates performance at modality-speciic behaviors.
Proc. Natl Acad. Sci. USA 104, 4153–4158 (2007).
93. Alvarez, I. etal. Heritable functional architecture in human visual cortex. Neuroimage
239, 118286 (2021).
94. Whitton, A. E., Treadway, M. T. & Pizzagalli, D. A. Reward processing dysfunction in major
depression, bipolar disorder and schizophrenia. Curr. Opin. Psychiatry 28, 7 (2015).
95. Lynch, C. J. etal. Automated optimization of TMS coil placement for personalized
functional network engagement. Neuron https://doi.org/10.1016/j.neuron.2022.08.012
(2022).
96. Cash, R. F. H. & Zalesky, A. Personalized and circuit-based transcranial magnetic
stimulation: evidence, controversies, and opportunities. Biol. Psychiatry https://doi.
org/10.1016/j.biopsych.2023.11.013 (2023).
97. Yeo, B. T. T. etal. The organization of the human cerebral cortex estimated by intrinsic
functional connectivity. J. Neurophysiol. 106, 1125–1165 https://doi.org/10.1152/
jn.00338.2011 (2011).
98. Kwon, Y. etal. Situating the parietal memory network in the context of multiple parallel
distributed networks using high-resolution functional connectivity. Preprint at bioRxiv
https://doi.org/10.1101/2023.08.16.553585 (2023).
99. Uddin, L. Q. etal. Controversies and progress on standardization of large-scale brain
network nomenclature. Netw. Neurosci 7, 864–905 (2023).
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in
published maps and institutional afiliations.
Open Access This article is licensed under a Creative Commons Attribution-
NonCommercial-NoDerivatives 4.0 International License, which permits any
non-commercial use, sharing, distribution and reproduction in any medium or
format, as long as you give appropriate credit to the original author(s) and the source, provide
a link to the Creative Commons licence, and indicate if you modiied the licensed material. You
do not have permission under this licence to share adapted material derived from this article
or parts of it. The images or other third party material in this article are included in the article’s
Creative Commons licence, unless indicated otherwise in a credit line to the material. If
material is not included in the article’s Creative Commons licence and your intended use is not
permitted by statutory regulation or exceeds the permitted use, you will need to obtain
permission directly from the copyright holder. To view a copy of this licence, visit http://
creativecommons.org/licenses/by-nc-nd/4.0/.
© The Author(s) 2024
Content courtesy of Springer Nature, terms of use apply. Rights reserved

Article
Methods
Datasets
The datasets used in this paper are described briefly below, with more
demographic and clinical information provided in theSupplemen
-
tary Information. No statistical tests were used to predetermine the
sample sizes. Overall, the depression sample collectively consisted
of 141 individuals (mean age 40.71 ± 13.82 years, 56.7% female) with
a diagnosis of major depression (based on DSM-IV-TR criteria and
confirmed by the Mini-International Neuropsychiatric Interview
administered by a trained clinician) drawn from five sources—SIMD,
mean age 29.47 ± 8.28 years, 3 female (F)/3 male (M), Weill Cornell
rTMS 1 (conventional 6-week rTMS, mean age 40.89 ± 12.73 years,
27 F/21 M), Weill Cornell rTMS 2 (accelerated, 1-week rTMS, mean age
40.89 ± 12.73 years, 21 F/24 M), Stanford University rTMS (conventional
6-week rTMS, mean age 38.09 ± 12.77 years, 29 F/13 M) and Weill Cornell
late-onset depression datasets (mean age 66.60 ± 5.31 years, 5 F/0 M).
The healthy control sample collectively consisted of 37 healthy adults
(mean age 31.72 ± 7.08 years, 11 F) drawn from six sources—the Weill
Cornell multi-echo (mean age 33.42 ± 9.10 years, 0 F/7 M)18,41, MyCon-
nectome (a single 45-year-old male)
12
, Midnight Scan Club (MSC; mean
age 29.1 ± 3.3 years, 5 F/5 M)3, cast-induced plasticity (a single 27-year-old
male)
100
, natural scenes dataset (NSD; mean age 26.50 ± 4.24 years,
6 F/2 M)
101
and Eskalibur datasets (mean age 31.4 ± 5.4 years, 5 F/5 M)
102
.
We note that the MSC, MyConnectome and cast-induced plasticity
study were obtained online (https://openneuro.org/) in a preprocessed,
fully denoised and surface-registered format, and no further preproc-
essing or denoising was performed for the present study.
MRI acquisition
Serial imaging of major depression dataset. Data were acquired on
a Siemens Magnetom Prisma 3 T scanner at the Citigroup Biomedi-
cal Imaging Center of Weill Cornell medical campus using a Siemens
32-channel head coil. Two multi-echo, multi-band resting-state fMRI
scans were collected using a T2*-weighted echo-planar sequence cover-
ing the full brain (TR 1,355 ms; TE1 13.40 ms, TE2 31.11 ms, TE3 48.82 ms,
TE4 66.53 ms and TE5 84.24 ms; FOV 216 mm; flip angle 68° (the Ernst
angle for grey matter assuming a T1 value of 1,400 ms); 2.4 mm isotropic
voxels; 72 slices; AP phase encoding direction; in-plane acceleration fac-
tor 2; and multi-band acceleration factor 6) with 640 volumes acquired
per scan for a total acquisition time of 14 min and 27 s. Spin echo EPI
images with opposite phase encoding directions (AP and PA) but iden-
tical geometrical parameters and echo spacing were acquired before
each resting-state scan. Multi-echo T1-weighted (TR/TI 2,500/1,000 ms;
TE1 1.7 ms, TE2 3.6 ms, TE3 5.5 ms and TE4 7.4 ms; FOV 256 mm; flip angle
8° and 208 sagittal slices with a 0.8 mm slice thickness) and T2-weighted
anatomical images (TR 3,200 ms; TE 563 ms; FOV 256; flip angle 8° and
208 sagittal slices with a 0.8 mm slice thickness) were acquired at the
end of each session.
Weill Cornell rTMS 1 and 2 datasets. MRI data were acquired on a Sie-
mens Magnetom Prisma 3 T machine at the Citigroup Biomedical Imag-
ing Center of Weill Cornell medical campus using a Siemens 32-channel
head coil. Two multi-echo, multi-band resting-state fMRI scans were
collected at each study visit using a T2*-weighted echo-planar sequence
covering the full brain (TR 1,300 ms; TE1 12.60 ms, TE2 29.5 1 ms, TE3
46.42 ms and TE4 63.33 ms; FOV 216 mm; flip angle 67° (the Ernst angle
for grey matter assuming a T1 value of 1,400 ms); 2.5 mm isotropic voxels;
60 slices; AP phase encoding direction; in-plane acceleration factor 2;
and multi-band acceleration factor 4) with 650 volumes acquired per
scan for a total acquisition time of 14 min and 5 s. Spin echo EPI images
with opposite phase encoding directions (AP and PA) but identical
geometrical parameters and echo spacing were acquired before each
resting-state scan. Multi-echo T1-weighted (TR/TI 2,500/1,000 ms; TE1
1.7 ms, TE2 3.6 ms, TE3 5.5 ms and TE4 7.4 ms; FOV 256 mm; flip angle 8°
and 208 sagittal slices with a 0.8 mm slice thickness) and T2-weighted
anatomical images (TR 3,200 ms; TE 563 ms; FOV 256 mm; flip angle
8° and 208 sagittal slices with a 0.8 mm slice thickness) were acquired
at the end of each session.
Stanford University rTMS dataset. MRI data were acquired on a GE
SIGNA 3 T machine at the Center for Neurobiological Imaging on Stan-
ford University campus using a Nova Medical 32-channel head coil.
Four multi-echo, multi-band resting-state fMRI scans were collected
using a T2*-weighted echo-planar sequence covering the full brain
(TR 1,330 ms; TE1 13.7 ms, TE2, 31.60 ms, TE3 49.50 ms and TE4 67.40 ms;
flip angle 67° (the Ernst angle for grey matter assuming a T1 value of
1,400 ms); 3 mm isotropic voxels; 52 slices; AP phase encoding direction;
in-plane acceleration factor 2; and multi-band acceleration factor 4)
with 338 volumes acquired per scan for a total acquisition time of
7 min and 30 s. Spin echo EPI images with opposite phase encoding
directions (AP and PA) but identical geometrical parameters and echo
spacing were acquired before each resting-state scan. T1-weighted
and T2-weighted anatomical images were acquired at the end of each
session.
Weill Cornell late-onset depression dataset. MRI data were
acquired on a Siemens Magnetom Prisma 3 T machine at the Citigroup
Biomedical Imaging Center of Weill Cornell medical campus using
a Siemens 32-channel head coil. Two multi-echo, multi-band
resting-state fMRI scans were collected at each study visit using
a T2*-weighted echo-planar sequence covering the full brain (TR
1,300 ms; TE1 12.60 ms, TE2 29.51 ms, TE3 46.42 ms and TE4 63.33 ms;
FOV 216 mm; flip angle 67° (the Ernst angle for grey matter assuming
a T1 value of 1,400 ms); 2.5 mm isotropic voxels; 60 slices; AP phase
encoding direction; in-plane acceleration factor 2; and multi-band
acceleration factor 4) with 480 volumes acquired per scan for a total
acquisition time of 10 min and 38 s. Spin echo EPI images with oppo-
site phase encoding directions (AP and PA) but identical geometrical
parameters and echo spacing were acquired before each resting-state
scan. Multi-echo T1-weighted (TR/TI 2,500/1,000 ms; TE1 1.7 ms, TE2
3.6 ms, TE3 5.5 ms and TE4 7.4 ms; FOV 256 mm; flip angle 8° and 208
sagittal slices with a 0.8 mm slice thickness) and T2-weighted anatomi-
cal images (TR 3,200 ms; TE 563 ms; FOV 256 mm; flip angle 8° and 208
sagittal slices with a 0.8 mm slice thickness) were acquired at the end of
each session.
Anatomical preprocessing and cortical surface generation
Anatomical data were preprocessed and cortical surfaces generated
using the Human Connectome Project (HCP) PreFreeSurfer, FreeSurfer
and PostFreeSurfer pipelines (v.4.3).
Multi-echo fMRI preprocessing
Preprocessing of multi-echo data minimized spatial interpolation and
volumetric smoothing while preserving the alignment of echoes. The
single-band reference (SBR) images (one per echo) for each scan were
averaged. The resultant average SBR images were aligned, averaged,
co-registered to the ACPC-aligned T1-weighted anatomical image and
simultaneously corrected for spatial distortions using FSL topup and
epi_reg programs. Freesurfer bbregister algorithm was used to refine this
co-registration. For each scan, echoes were combined at each time point
and a unique 6 DOF registration (one per volume) to the average SBR
image was estimated using FSL MCFLIRT tool, using a four-stage (sinc)
optimization. All of these steps (co-registration to the average SBR
image, ACPC alignment and correcting for spatial distortions) were
concatenated using FSL convertwarp tool and applied as a single
spline warp to individual volumes of each echo after correcting for
slice time differences using FSL slicetimer program. The functional
images underwent a brain extraction using the co-registered brain
extracted T1-weighted anatomical image as a mask and corrected for
Content courtesy of Springer Nature, terms of use apply. Rights reserved

signal intensity inhomogeneities using ANT N4BiasFieldCorrection
tool. All denoising was performed on preprocessed, ACPC-aligned
images.
Multi-echo fMRI denoising
Preprocessed multi-echo data were submitted to multi-echo ICA
(ME-ICA103), which is designed to isolate spatially structured T2*-
dependent (neurobiological; BOLD-like) and S0-dependent
(non-neurobiological; not BOLD-like) signals and implemented using
the tedana.py workflow
104
. In short, the preprocessed, ACPC-aligned
echoes were first combined according to the average rate of T2* decay
at each voxel across all time points by fitting the monoexponential
decay, S(t) = S0e
−t/T2*
. From these T2* values, an optimally combined
multi-echo (OC-ME) time series was obtained by combining echoes
using a weighted average (WTE = TE × e
−TE/T2*
). The covariance structure
of all voxel time courses was used to identify main signals in the OC-ME
time series using principal component and independent component
analysis. Components were classified as either T2*-dependent (and
retained) or S0-dependent (and discarded), primarily according to their
decay properties across echoes. All component classifications were
manually reviewed by C.J.L. and revised when necessary following the
criteria described in ref. 105. Mean grey matter time-series regression
was performed to remove spatially diffuse noise. Temporal masks were
generated for censoring high-motion time points using a framewise
displacement threshold of 0.3 mm and a backward difference of two
TRs, for an effective sampling rate comparable to historical framewise
displacement measurements (approximately 2–4 s). Before the frame-
wise displacement calculation, head realignment parameters were
filtered using a stopband Butterworth filter (0.2–0.35 Hz) to attenuate
the influence of respiration106 on motion parameters.
Single-echo fMRI denoising
The following denoising procedures were applied to the NSD and ABCD
datasets. The NSD dataset was obtained in an already preprocessed (but
not yet denoised) format. For the ABCD data, Fast Track (unprocessed)
neuroimaging data were obtained by means of NDA command line
utilities (https://github.com/NDAR/nda-tools) and subjected to the
preprocessing steps used for multi-echo fMRI data (omitting steps
involving combination of echoes). Preprocessed single-echo data were
then submitted to ICA-AROMA. All component classifications were
manually reviewed by C.J.L. and revised when necessary following
the criteria described in ref. 105. Mean grey matter time-series regres-
sion was performed to remove spatially diffuse noise. Temporal masks
were generated for censoring high-motion time points, as done for the
multi-echo fMRI datasets.
Surface processing and CIFTI generation of fMRI data
The denoised fMRI time series was mapped to the individual’s fsLR 32k
midthickness surfaces with native cortical geometry preserved (using
the -ribbon-constrained method), combined into the connectivity
informatics technology initiative (CIFTI) format and spatially smoothed
with geodesic (for surface data) and Euclidean (for volumetric data)
Gaussian kernels (σ = 2.55 mm) using Connectome Workbench com-
mand line utilities
107
. This yielded time courses representative of the
entire cortical surface, subcortex (accumbens, amygdala, caudate,
hippocampus, pallidum, putamen, thalamus and brainstem) and
cerebellum but excluding non-grey matter tissue. Spurious coupling
between subcortical voxels and adjacent cortical tissue was mitigated
by regressing the average time series of cortical tissue of less than
20 mm in Euclidean space from a subcortical voxel.
Precision mapping of functional brain networks in individuals
A functional connectivity matrix summarizing the correlation between
the time courses of all cortical vertices and subcortical voxels across
all study visits was constructed. Correlations between nodes 10 mm
or less apart (geodesic and Euclidean space used for cortico–cortical
and subcortical–cortical distance, respectively) were set to zero. Cor-
relations between voxels belonging to subcortical structures were set
to zero. Functional connectivity matrices were thresholded in such
a way that they retained at least the strongest X% correlations (0.01%,
0.02%, 0.05%, 0.1%, 0.2%, 0.5%, 1%, 2% and 5%) to each vertex and voxel
and were used as inputs for the InfoMap community detection algo-
rithm
108
, one of the most widely used approaches for delineating func-
tional brain networks and their boundaries in individuals
2,3,12,18,41,45
. Free
parameters (for example, the number of algorithm repetitions) for the
Infomap algorithm were fixed across subjects. The total number of
communities identified by Infomap is controlled in part by how many
connections are retained in the functional connectivity matrix after
thresholding. The optimal scale for further analysis across individuals
was defined as the graph threshold producing the best size-weighted
average homogeneity relative to the median of the size-weighted aver-
age homogeneity calculated from randomly rotated networks, as done
in ref. 109. Size-weighted average homogeneity was maximized rela-
tive to randomly rotated communities at the 0.1% graph density and
resulted in 89.13 ± 8.04 communities on average across individuals.
Each Infomap community was algorithmically assigned to one
of 20 possible functional network identities (Default-Parietal,
Default-Anterolateral, Default-Dorsolateral, Default-Retrosplenial,
Visual-Lateral, Visual-Stream, Visual-V1, Visual-V5, Frontoparietal,
Dorsal Attention, Premotor/Dorsal Attention II, Language, Sali-
ence, Cingulo-opercular/Action-mode
41
, Parietal memory, Auditory,
Somatomotor-Hand, Somatomotor-Face, Somatomotor-Foot, Audi-
tory or Somato-Cognitive-Action) primarily according their functional
connectivity and spatial locations relative to a specified set of priors.
All algorithmic assignments were manually reviewed by C.J.L. and
manually adjusted in the case of an ambiguous assignment. See Sup-
plementary Fig.14 for more details about algorithmic assignments and
Supplementary Figs.15 and 16 for the visualizations of the functional
network priors used in this study.
Functional brain networks were also mapped brain-wide using the
multiplex version of the InfoMap community detection algorithm
110
. In
a multiplex network, physical nodes (brain regions) can exist in several
layers (study visits). A temporal network (node × node × study visit)
summarizing the correlation between the time courses of all cortical
vertices and subcortical voxels across study visits was constructed for
each patient. Correlations between nodes less than 10 mm apart (geo-
desic and Euclidean space used for cortico–cortical and subcortical–
cortical distance, respectively) were set to zero. Correlations between
voxels belonging to subcortical structures were set to zero. Links
between layers were generated automatically using neighbourhood
flow coupling. The temporal distance between layers was constrained to
1 using the --multilayer-relax-limit option to encode the temporal order
of study visits. Multiplex functional network parcellations were used
for the analyses performed in the section ‘Salience network topology
is trait-like’ and Fig.3a–e.
Calculating functional network size and spatial locations in
individuals
We first measured the surface area (in mm
2
) that each vertex in the
individual’s midthickness surface is responsible for (wb_command
--surface-vertex-areas). Next, we calculated the relative contribution
(size) of each functional network to the total cortical surface area by
taking the total surface area of all network vertices in relation to the
total cortical surface area. In the striatum, in which each voxel rep-
resents the same amount of tissue, the relative contribution of each
functional network to the total striatal volume was calculated by taking
the total number of network voxels in relation to the total striatal voxels.
The statistical significances of group differences in network size were
evaluated using permutation tests and independent sample t-tests
(the latter implemented using Matlab ttest2.m function). Effect size
Content courtesy of Springer Nature, terms of use apply. Rights reserved

Article
(Cohen’s d) was calculated as the difference in group means divided
by pooled standard deviation. Assumptions about equal variance were
adjusted when appropriate (based on two-sample F-tests performed
using Matlab vartest2.m function). The relative difference between
groups was calculated as the absolute difference divided by network
size in healthy controls. Density maps were created by calculating the
percentage of individuals with salience network representation at
each cortical vertex or striatum voxel. These procedures collectively
correspond to the analyses performed in the section ‘Connectivity
state predicts anhedonia’ and Fig.1c–e.
Classification analysis
Functional network size (the percentage of total cortical surface area
occupied by each network, 20 networks/features total) were used as
predictive features in a support vector machine classifier to distin-
guish individuals with depression and healthy controls. The model
was trained using repeated (100 iterations) nested split-half (twofold)
cross-validation with a grid search optimization strategy for hyperpa-
rameter tuning (box constraint and kernel size). The synthetic minority
oversampling technique (SMOTE
111
) was used to prevent classifica-
tion bias in favour of the majority class and was performed on training
data only to prevent data leakage. Classification accuracy was calcu-
lated as the percentage of correct predictions and statistical signifi-
cance assessed using permutation tests (shuffled diagnostic labels
and 1,000 iterations). A confusion matrix was created using Matlab
confmat.m function. Feature importance was evaluated by iteratively
omitting each functional network and evaluating the resulting loss in
accuracy. These procedures collectively are related to the analyses
performed in the section ‘Salience network expansion in depression’
and Fig.1f–i.
Evaluating how salience network expansion displaces other
functional systems
The parts of each depressed individual’s salience network map that
did and did not overlap with the salience network in the group-average
healthy control map were operationalized as ‘non-encroaching’ and
‘encroaching’, respectively. The group-average healthy control map
was obtained by calculating the mode assignment across healthy con-
trols at each point in the brain. Encroaching clusters were identified
(wb_command -cifti-find-clusters) and were classified as border shifts
if any part of the cluster was within 3.5 mm (in geodesic space) of a sali-
ence network vertex in the group-average healthy control map, and as
ectopic intrusions if they did not, as done in ref. 41. An encroachment
profile was calculated as the relative contribution of each functional
network to the total surface area of the encroaching portion of the sali-
ence network. Individuals were clustered on the basis of the similarity
of their encroachment profiles using the Louvain method (commu-
nity_louvain.m function from the Brain Connectivity Toolbox
112
). These
procedures correspond to the analyses performed in the section ‘Three
salience network expansion modes’ and Fig.2a–g.
Assessing the stability of salience network topography across
time
The multiplex versions of each individual’s salience network were
used to assess the extent to which network topography (size) varied
across study time points in highly sampled individuals with and without
depression. Variability in salience network size was correlated with
the overall severity of depressive symptoms (HDRS6) using Matlab
corr.m function. In the replication samples (Weill Cornell Medicine
rTMS 1, Weill Cornell Medicine rTMS 2 and Stanford University rTMS
samples), we assessed pre-to-post change in salience network topog-
raphy using paired two-tailed paired sample t-tests by means of Matlab
ttest.m function. Data were binned according to treatment duration
(conventional 6-week rTMS or accelerated 1-week rTMS). The number
of depressive episodes in each individual’s lifetime was inferred from
their Mini-International Neuropsychiatric Interview. These proce-
dures collectively correspond to the analyses performed in the section
‘Salience network topology is trait-like’ and Fig.3.
Evaluating salience network topography early in life before
symptom onset
We used the ABCD dataset (release 5.0) to test if atypical salience net-
work topology precedes the onset of depression symptoms. Symp-
toms of depression in the ABCD study were operationalized using the
ASEBA DSM-oriented scale for depression (mh_p_cbcl.csv) from the
ABCD parent child behaviour checklist (CBCL). After excluding sub-
jects with missing behavioural data or those with MRI data flagged
internally ABCD for data quality issues, we identified n = 58 subjects
(37 F) meeting criteria for onset of clinical depression symptoms at the
3-year follow-up (t-score 70 or more at or after the 3-year follow-up and
t-scores below 65 at the previous study visits). One participant’s data
were not accessible on Fast Track, resulting in n = 57 total. An equal
number of subjects with no clinically significant depression symptoms
at any study time point (t-scores more than 65 at all study time points)
were randomly selected as a control sample. The statistical significance
of group differences in salience network size were evaluated using
permutation tests and independent sample t-tests (the latter imple-
mented using Matlab ttest2.m function). Assumptions about equal
variance were adjusted when appropriate (based on two-sample F-tests
performed using Matlab vartest2.m function). These procedures col-
lectively correspond to the analyses performed in the section ‘Salience
network topology is trait-like’ and Fig.3g.
Longitudinal analyses relating changes in connectivity with
symptom severity
Composite measures of anhedonia- and anxiety-related symptoms were
obtained instructing three clinicians (I.E., J.D.P. and N.S.) to quantify
(on a scale of 0–3; 0, not at all; 1, somewhat; 2, largely; 3, very strongly)
the extent to each item from the battery of clinical scales administered
to the SIMD subjects reflects anhedonia- or anxiety-related symptoms.
Items assigned a score of 1 or greater by all three clinicians were included
in the composite measures (Supplementary Fig.9). For each subject,
separately, the consensus items were mininum–maximum normalized,
adjusted for valence (so that higher scores reflect more severe symp-
toms across all items) and then subjected to a principal component
analysis to extract a time course (PC1) of anhedonia or anxiety severity
across study visits. To validate this approach, we quantified the similar-
ity to validated measures of anhedonia and anxiety using independent
data and observed good correspondence (Pearson correlations more
than 0.4). Functional connectivity strength between all pairs of corti-
cal (anterior cingulate, lateral prefrontal and anterior insula cortex)
and striatal (nucleus accumbens, caudate and putamen) nodes of the
salience network was calculated for each study visit, separately, and
correlated with the anhedonia or anxiety PC1. This analysis was con-
strained to the three major cortical and striatal nodes of the salience
network in part to reduce the likelihood of false positives. Correlations
not exceeding chance (based on null distribution of correlation coef-
ficients obtained using rotated clinical data) were set to zero. Circular
permutation tests (using Matlab circshift.m function) were used to
preserve temporal autocorrelation. Cross-correlation analyses were
performed using Matlab crosscorr.m function (with NumLags set to 2).
For the cross-sectional analysis, the total Snaith–Hamilton pleasure
scale (SHAPS
113
) score was calculated using baseline clinical data. These
procedures collectively correspond to the analyses performed in the
section ‘Connectivity state predicts anhedonia’ and Fig.4.
Clinical trial information
A portion of the data used in this report was obtained from the
biomarker-guided rTMS for treatment-resistant depression study
(NCT04041479), randomized controlled trial of conventional versus
Content courtesy of Springer Nature, terms of use apply. Rights reserved

theta burst rTMS (HFL versus TBS) (NCT01887782) and accelerated
TMS for depression and obsessive compulsive disorder studies
(NCT04982757).
Reporting summary
Further information on research design is available in theNature Port-
folio Reporting Summary linked to this article.
Data availability
An example subject (HC-1) from the Weill Cornell Multi-echo dataset
is available in an OpenNeuro repository at https://openneuro.org/
datasets/ds005118/versions/1.0.0. Other data from the Weill Cornell
Multi-echo and Eskalibur datasets are available on request from the
corresponding authors of their respective publications, pending
institutional approval of a data-sharing agreement in compliance
with their respective IRB protocols. Data from the MyConnectome
dataset are available on OpenNeuro repository at https://openneuro.
org/datasets/ds000031/versions/2.0.2. Data from the MSC dataset are
available in the OpenNeuro repository at https://openneuro.org/data-
sets/ds000224/versions/1.0.4. Data from the cast-induced plasticity
dataset is available in the OpenNeuro repository at https://openneuro.
org/datasets/ds002766/versions/3.0.0. Data from the natural scenes
dataset are available from Amazon Web Services at https://registry.
opendata.aws/nsd/. The ABCD data used in this report are from Annual
Release (5.0). Data from individual subjects with depression are a part
of ongoing clinical trials and not publicly available now but will be made
available after completion of the trials through the NIMH Data Archive
and other clinical data repositories.
Code availability
Code for preprocessing multi-echo fMRI data is maintained in an
online repository (https://github.com/cjl2007/Liston-Laboratory-
MultiEchofMRI-Pipeline). Code for performing the precision functional
mapping and network size calculations described in this manuscript
are maintained in an online repository (https://github.com/cjl2007/
PFM-Depression). Software packages incorporated into the above
pipelines for data analysis included: Matlab R2019a, https://www.math-
works.com/; Connectome Workbench 1.4.2, http://www.humancon-
nectome.org/software/connectome-workbench.html; Freesurfer v6,
https://surfer.nmr.mgh.harvard.edu/; FSL 6.0, https://fsl.fmrib.ox.ac.
uk/fsl/fslwiki; and Infomap, https://www.mapequation.org.
100. Newbold, D. J. etal. Plasticity and spontaneous activity pulses in disused human brain
circuits. Neuron 107, 580–589 (2020).
101. Allen, E. J. etal. A massive 7T fMRI dataset to bridge cognitive neuroscience and artiicial
intelligence. Nat. Neurosci. 25, 116–126 (2022).
102. Moia, S. etal. ICA-based denoising strategies in breath-hold induced cerebrovascular
reactivity mapping with multi echo BOLD fMRI. Neuroimage 233, 117914 (2021).
103. Kundu, P. etal. Integrated strategy for improving functional connectivity mapping using
multiecho fMRI. Proc. Natl Acad. Sci. USA 110, 16187–16192 (2013).
104. DuPre, E. etal. TE-dependent analysis of multi-echo fMRI with tedana. J. Open Source
Softw. 6, 3669 (2021).
105. Griffanti, L. etal. Hand classiication of fMRI ICA noise components. Neuroimage 154,
188–205 (2017).
106. Power, J. D. etal. Distinctions among real and apparent respiratory motions in human fMRI
data. Neuroimage 201, 116041 (2019).
107. Marcus, D. S. etal. Human Connectome Project informatics: quality control, database
services, and data visualization. Neuroimage 80, 202–219 (2013).
108. Rosvall, M. & Bergstrom, C. T. Maps of random walks on complex networks reveal
community structure. Proc. Natl Acad. Sci. USA 105, 1118–1123 (2008).
109. Gordon, E. M. etal. Default-mode network streams for coupling to language and control
systems. Proc. Natl Acad. Sci. USA 117, 17308–17319 (2020).
110. De Domenico, M., Lancichinetti, A., Arenas, A. & Rosvall, M. Identifying modular lows on
multilayer networks reveals highly overlapping organization in interconnected systems.
Phys. Rev. X 5, 011027 (2015).
111. Chawla, N. V., Bowyer, K. W., Hall, L. O. & Kegelmeyer, W. P. SMOTE: synthetic minority
over-sampling technique. J. Artif. Intell. Res. 16, 321–357 (2002).
112. Rubinov, M. & Sporns, O. Complex network measures of brain connectivity: uses and
interpretations. Neuroimage 52, 1059–1069 (2010).
113. Snaith, R. P. etal. A scale for the assessment of hedonic tone the Snaith–Hamilton
pleasure scale. Br. J. Psychiatry 167, 99–103 (1995).
Acknowledgements We thank the staff at the Citigroup Biomedical Imaging Center for
assistance with data collection. We thank all study participants and especially the SIMD
study participants for their time and dedication to science. R. Kong and T. Yeo helped with
implementing the multi-session hierarchical Bayesian modelling42 method. This work was
supported by grants to C.L. from the National Institute of Mental Health, the National Institute
on Drug Addiction, the Hope for Depression Research Foundation and the Foundation
for OCD Research. C.J.L. was supported by an NIMH F32 National Research Service Award
(F32MH120989). N.S. was supported by K23 MH123864. Work on ‘Personalized therapeutic
neuromodulation for anhedonic depression’ is supported by Wellcome Leap as part of the
Multi-Channel Psych Program.M.L. was supported by Deutsche Forschungsgemeinschaft
(DFG grant CRC 1193, subproject B01).
Author contributions C.J.L., I.E., J.D.P. and C.L. conceived this work. Data acquisition, analysis
and interpretation were conducted by C.J.L., I.E., T.N., A. Ayaz, S.Z., D.W., N.M., M.J., M.C., J.C.,
I.S., C.H., M.L., H.B., D.B., L.W.V., N.S., E.G., S.M., C.C.-G., J.D., F.V.-R., Z.J.D., D.M.B., K.K., A. Aloysi,
E.M.G., M.T.B., N.W., J.D.P., B.Z., L.G., F.M.G. and C.L. Manuscript writing and revision was done
by C.J.L., I.E., E.M.G., J.D.P. and C.L. Study authors C.J.L., I.E., M.L. and J.D.P. were included as
participants.
Competing interests C.L. and C.J.L. are listed as inventors for Cornell University patent
applications on neuroimaging biomarkers for depression which are pending or in preparation.
C.L. has served as a scientiic advisor or consultant to Compass Pathways PLC, Delix
Therapeutics and Brainify.AI. The remaining authors declare no competing interests.
Additional information
Supplementary information The online version contains supplementary material available at
https://doi.org/10.1038/s41586-024-07805-2.
Correspondence and requests for materials should be addressed to Charles J. Lynch or
Conor Liston.
Peer review information Nature thanks Avram Holmes, Theodore Satterthwaite, Chao-Gan Yan
and the other, anonymous, reviewer(s) for their contribution to the peer review of this work.
Reprints and permissions information is available at http://www.nature.com/reprints.
Content courtesy of Springer Nature, terms of use apply. Rights reserved

Article
Extende d Data Fig. 1 | Se rial Imagi ng of Major De pression . a, The SIMD
project invol ved repeated mu lti-echo rest ing-state f MRI scans (M E-rsfMRI) a nd
clinical as sessment s of six individua ls with depres sion over long peri ods of
time. Pre cision funct ional mapping wa s then used to 1) inves tigate differ ences
in functi onal network top ology, specif ically size relat ive to healthy cont rols,
and 2) identif y which atypi cal aspect s of network top ology or conne ctivity a re
stable vers us sensitive to mo od state wit hin individual s as the severity o f their
symptoms f luctua ted, and they cycle d in and out of depres sive episode s.
Images cre ated with BioRender.com. b, The relative cont ribution (size) of eac h
functio nal network to th e total cortic al surface are a was obtained by t aking the
total surf ace area of all net work vertices in r elation to the tot al cortical su rface
area. Thi s approach control s for fact that eac h cortical ver tex represen ts a
different a mount of surfac e area (SA). In the striat um, where each voxel
represen ts the same amo unt of tissue, the r elative contri bution of each
functio nal network to th e total striat al volume was calcul ated by taking t he
total numb er of network voxels in re lation to the tota l striatal voxels.
Content courtesy of Springer Nature, terms of use apply. Rights reserved

Extende d Data Fig. 2 | Sal ience net work expansion i n depress ion remains
statistically significant when controlling for sex ratio imbalance, and
individual differences in head motion and age. a, Salience n etwork size was
regress ed against se x (a variable of non-inter est that differ s between th e two
groups) and grou p comparison s were repeated u sing the residu als (e). The
salience n etwork was still s ignific ant larger in the Se rial Imaging o f Major
Depress ion (SIMD; two-t ailed indepen dent sample t-test, P < 0.0 01, T = 7.02, and
Cohen’s d = 2.09) and in all three replica tion samples (two -tailed inde pendent
sample t-tests, a ll P < 0.001, T’s > 3.00, and Cohen’s d > 0.6) relative to he althy
controls. b-c, This analysis was re peated when a lso including he ad motion
(operational ized as the % of volume r etained af ter motion cen soring) and age
(in years) as additi onal covariates . In all of these mo dels, the salie nce network
remained significantly larger in the SIMD (two-tailed, independe nt sample
t-te st , P < 0.001, T = 6.75, and Cohen’s d = 2.06) and in al l three replicat ion
samples (two-tailed independent sample t-tests, all P’s ≤ 0.002, T’s > 2.2, and
Cohen’s d > 0.56) rela tive to healthy con trols. All erro r bars represen t standard
deviation.
Content courtesy of Springer Nature, terms of use apply. Rights reserved

Article
Extende d Data Fig. 3 | Exp ansion of the s alience ne twork accompa nied by
contraction of neighboring functional systems. a-b, The salience (SA L,
black), default mo de (DMN, red), fronto parietal (FP, yellow), and cingulo-
opercular (CO, purpl e) networks in a grou p-average map of heal thy controls
versus 3 repre sentative ind ividuals with d epression. E xpansion of the s alience
network in co rtex (see Fig.1c) was acco mpanied in some c ases by contr action
of other fun ctional net works — most not ably the cingulo -opercular ne twork
(two-tailed p ermutati on test, *P = 0.04, uncor rected, Z-score = 2 .09, n = 43), but
this effec t was not obser ved in any of the repli cation sampl es. All error ba rs
represent standard deviation .
Content courtesy of Springer Nature, terms of use apply. Rights reserved

Extende d Data Fig. 4 | Evid ence of salie nce network exp ansion in lar ge
n group-average datasets. a, Salience network m apped using t wo large
n group-average d ata from previo us studies of he althy controls oc cupy 1.27%
and 1.98% of cor tex. The group -average HCP funct ional connec tivity mat rix
(which only inclu des subject s with restin g-state fM RI data recon structed w ith
the r227 re con algorithm) was o btained from t he S1200 rele ase and subjec ted
to the same pre cision func tional mapping pr ocedures appl ied to individual
subject s in the main text. T he WU120 s alience net work map was obta ined
online (https://balsa.wustl.edu/jNXKl). b, Salie nce network map ped using
large n group-aver age data and previ ous studies of de pression occ upies
betwee n 3.28% (mode as signment of all i ndividuals wi th depression i n current
study) and 3.43% of tot al cortica l surface area. G roup-averaged fun ctional
connec tivity was c alculated in the T HREE-D sample us ing group-level PC A
(MELODIC Inc remental Grou p-PCA, MIG P), and the resultant g roup-average
FC matrix was s ubjected to t he same precisi on functiona l mapping proce dures
applied to ind ividual subje cts in the main tex t.
Content courtesy of Springer Nature, terms of use apply. Rights reserved

Article
Extende d Data Fig. 5 | Wit hin-perso n stabilit y of salience n etwork topol ogy
and connectivity. a-b, Split-half reliability test ing of salience n etwork
topolog y and functio nal connecti vity in the le ast (SIMD-2, 58 min of f MRI
scanning t otal) and most (SIM D-4, 29.96 hrs. of fMR I scanning to tal) sampled
individual s with depres sion from the Ser ial Imaging of M ajor Depressi on
(SIMD) dataset.
Content courtesy of Springer Nature, terms of use apply. Rights reserved

Extende d Data Fig. 6 | Se e next page for capti on.
Content courtesy of Springer Nature, terms of use apply. Rights reserved

Article
Extende d Data Fig. 6 | Sal ience net work expansion i n depressi on
disproportionately affects heteromodal systems neighboring it, not
unimodal sensorimotor networks. a, On average across subj ects, the maj ority
of salienc e network expans ion in depressi on affected e ither the Defa ult-
parietal, Frontoparietal, or Cingulo-op ercular networks. In contrast, Salienc e
network encroachment upon unimodal sensorimotor net works (for example,
the visual, auditory, somatomotor subnetworks) was absent. b, The average
map of Salien ce network enc roachment was c ompared to 73 can onical maps of
the brain’s functional and structural architecture (“annotations”) obtained
from the neur omaps toolbox 53 to help identify possible biological mechanisms
for its expans ion in individual s with depres sion. Thes e maps are derived f rom
a variety of independent molecular, microstructural, electrophysiological,
developmen tal, and funct ional datas ets. Spatial s imilarity was q uantifie d using
Spearman rank correlation, and statistical significance evaluated via spatial
autocorrelation preserving null models. We observed multiple significant
associations — including with principal gradients of functional connectivity
and gene expre ssion, and the sp atial distrib ution of neurotr ansmitter re ceptors
(μ-opioid, histamine H3 receptors), intracortical myelin, and individual
variability in functional connectivity.
Content courtesy of Springer Nature, terms of use apply. Rights reserved

Extende d Data Fig. 7 | Se e next page for capti on.
Content courtesy of Springer Nature, terms of use apply. Rights reserved

Article
Extende d Data Fig. 7 | Sa lience net work expansi on in depres sion is as sociated
with stable patterns of atypical functional connectivity. a-b, Evaluating
functional connectivity strength between encroaching and non-encroaching
vertice s of the Salience n etwork relative to r unner-up network as signments
network . Strength of fu nctional con nectivit y between e ncroaching nod es of
the salien ce network and th e rest of the Salie nce network, a nd the functio nal
network s that typic ally occupy that spa ce in healthy cont rols (most often
Default, Frontoparietal, or Cingulo-opercular). This analysis was per formed
using split ha lves of each indiv idual’s resting-st ate fMR I dataset to eva luate the
stabilit y of the Salienc e network assi gnment asso ciated with th e “encroaching ”
vertice s relative to the runn er-up assignment s. Functional c onnectiv ity
betwee n encroaching S alience netwo rk vertices an d the rest of the Sali ence
network wa s on average 59% stronge r than with the run ner-up network (two-
tailed ind ependent sa mple t-test, all P’s < 0.001, Bonferron i correction , n = 141).
This was th e case when usin g either the f irst (a) or second (b) half of e ach
individual ’s concatenate d resting-sta te fMRI da taset, indic ating good st ability.
Error bars rep resent stan dard deviation.
Content courtesy of Springer Nature, terms of use apply. Rights reserved

Extende d Data Fig. 8 | In creased cor tical repr esentat ion of salien ce
network in adults with late-onset depression. Five individual s (mean age =
66.60 ± 5. 31 years, 5 F) with a di agnosis of majo r depression an d met criteria for
late-ons et depressio n (LOD, defined he re as onset of f irst depress ive episode at
or after th e age of 60) underwent r epeated clini cal assess ments and f MRI scans
(6 × 10.64 min multi- echo resting-s tate fMR I scans, 63. 84 min total p er-subject)
before, durin g, and after a br ief evidence -based psychot herapy. Salience
network wa s larger in these in dividuals wit h LOD relative to healt hy controls
(two-tailed p ermutati on test, *P = 0.009, uncorre cted, Z-score = 2.90). The
n = 37 healthy con trol data are also sh own in the main text F ig.1c. Error bars
represent standard deviation .
Content courtesy of Springer Nature, terms of use apply. Rights reserved

Article
Extende d Data Fig. 9 | De nse-sam pling of depr essive symp toms and
functional connectivity in a second individual with depression. a, A heat
map summari zes fluc tuations in ind ividual items s elected fro m a variety of
clinical int erviews and s elf-report scale s related to anhe donia in an example
individual (SI MD-6). Clinical da ta was resampl ed (using shape- pre serving
piecewi se cubic interp olation) to days for visu alization purp oses (black a nd
red dots ab ove heat map mark the dat es of study visi ts and ECT treatm ents
received unr elated to the pres ent study, respec tively). b, Functional
connec tivity of sali ence network voxels i n nucleus accum bens (NAc) when
symptoms o f anhedonia are low (stu dy visits in the b ottom quart ile) and high
(study visit s in top quartile). c, Bootstrap resampling (iteratively selecting
50% of all time p oints at random , and logging c orrelation be tween nucleu s
accumbe ns ←→ anterior ci ngulate FC and anhe donia) indicated go od stabilit y.
Content courtesy of Springer Nature, terms of use apply. Rights reserved

Extende d Data Fig. 10 | Lon g term asse ssment of an hedonia an d anxiety
related symptoms in two deeply-sampled individuals with major
depression. a-b, Heat map summar izes flu ctuations in i ndividual item s related
to anxiet y (blue, 27 item s total) that were se lected from a va riety of clinic al
interv iews and self-repor t scales comp leted by two dee ply-sampled in dividuals
with depre ssion (a, SIMD- 4; b, SIMD-6). Head maps for a nhedonia relate d items
are shown in mai n text Fig.4a and Extend ed Data Fig.9a for SIM D-4 and
SIMD- 6, respect ively. Clinical data was re sampled (using sh ape-preser ving
piecewi se cubic interp olation) to days for visua lization purp oses. The f irst
principal c omponent (PC1) o f the anhedonia an d anxiety me asures were
modestl y correlated wi th one another w ithin each indiv idual over time
(Pearson co rrelation, MDD 04: r = 0.41, P < 0.001; MDD06: r = 0.45, P < 0.001),
indicating that the severity of anxiety and anhedonia related symptoms can
fluc tuate indepe ndently of one ano ther, but also that they bo th respond to
global shif ts in illness s everity, which were prima rily related to ECT in SI MD-6).
Content courtesy of Springer Nature, terms of use apply. Rights reserved

1
nature
portfolio
|
reporting
summary
April
2023
Corresponding
author(s):
Last
updated
by
by
author(s):
Reporting
Summary
Nature
Portfolio
wishes
to
to
improve
the
reproducibility
of
of
the
work
that
we
we
publish.
This
form
provides
structure
for
consistency
and
transparency
in
in
reporting.
For
further
information
on
on
Nature
Portfolio
policies,
see
our
Editorial
Policies
and
the
Editorial
Policy
Checklist
Statistics
For
all
statistical
analyses,
confirm
that
the
following
items
are
present
in
in
the
figure
legend,
table
legend,
main
text,
or
or
Methods
section.
n/a
Confirmed
The
exact
sample
size
(
n
)
for
each
experimental
group/condition,
given
as
as
a
discrete
number
and
unit
of
of
measurement
A
statement
on
on
whether
measurements
were
taken
from
distinct
samples
or
or
whether
the
same
sample
was
measured
repeatedly
The
statistical
test(s)
used
AND
whether
they
are
one-
or
or
two-sided
Only common tests should be described solely by name; describe more complex techniques in the Methods section.
A
description
of
of
all
covariates
tested
A
description
of
of
any
assumptions
or
or
corrections,
such
as
as
tests
of
of
normality
and
adjustment
for
multiple
comparisons
A
full
description
of
of
the
statistical
parameters
including
central
tendency
(e.g.
means)
or
or
other
basic
estimates
(e.g.
regression
coefficient)
AND
variation
(e.g.
standard
deviation)
or
or
associated
estimates
of
of
uncertainty
(e.g.
confidence
intervals)
For
null
hypothesis
testing,
the
test
statistic
(e.g.
F
,
t
,
r
)
with
confidence
intervals,
effect
sizes,
degrees
of
of
freedom
and
P
value
noted
Give P values as exact values whenever suitable.
For
Bayesian
analysis,
information
on
on
the
choice
of
of
priors
and
Markov
chain
Monte
Carlo
settings
For
hierarchical
and
complex
designs,
identification
of
of
the
appropriate
level
for
tests
and
full
reporting
of
of
outcomes
Estimates
of
of
effect
sizes
(e.g.
Cohen's
d
,
Pearson's
r
),
),
indicating
how
they
were
calculated
Our web collection on
statistics for biologists
contains articles on many of the points above.
Software
and
code
Policy
information
about
availability
of
of
computer
code
Data
collection
Data
analysis
For
manuscripts
utilizing
custom
algorithms
or
or
software
that
are
central
to
to
the
research
but
not
yet
described
in
published
literature,
software
must
be
be
made
available
to
to
editors
and
reviewers.
We
We
strongly
encourage
code
deposition
in
in
a
community
repository
(e.g.
GitHub).
See
the
Nature
Portfolio
guidelines
for
submitting
code
&
software
for
further
information.
Charles
Lynch
Ph.D.
Conor
Liston
M.D.,
Ph.D.
5/30/2024
No
No
software
was
used
for
data
collection.
Code
for
preprocessing
multi-echo
fMRI
data
is
is
maintained
in
in
an
an
online
repository
(https://github.com/cjl2007/Liston-Laboratory-
MultiEchofMRI-Pipeline).
Code
for
performing
precision
functional
mapping
and
code
specific
to
to
the
analyses
performed
in
in
this
manuscript
are
maintained
in
in
an
an
online
repository
(https://github.com/cjl2007/PFM-Depression).
Software
packages
incorporated
into
the
above
pipelines
for
data
analysis
included:
Matlab
R2019a,
https://www.mathworks.com/;
Connectome
Workbench
1.4.2,
http://www.humanconnectome.org/software/connectome-workbench.html;
Freesurfer
v6,
https://
surfer.nmr.mgh.harvard.edu/;
FSL
6.0,
https://fsl.fmrib.ox.ac.uk/fsl/fslwiki;
and
Infomap
v2.0.0,
https://www.mapequation.org.
Advanced
Normalization
Tools
(ANTS;
v2.3.4).
Content courtesy of Springer Nature, terms of use apply. Rights reserved

2
nature
portfolio
|
reporting
summary
April
2023
Data
Policy information about availability of data
All manuscripts must include a data availability statement. This statement should provide the following information, where applicable:
- Accession codes, unique identifiers, or web links for publicly available datasets
- A description of any restrictions on data availability
- For clinical datasets or third party data, please ensure that the statement adheres to our policy
Research involving human participants, their data, or biological material
Policy information about studies with human participants or human data. See also policy information about sex, gender (identity/presentation),
and sexual orientation and race, ethnicity and racism.
Reporting on sex and gender
Reporting on race, ethnicity, or
other socially relevant groupings
Population characteristics
Recruitment
Data from the Weill Cornell Multi-echo and Eskalibur datasets are available on reasonable request from C.J.L, I.E., J.D.P, C.L., and S.M., C.C.
Data from the MyConnectome dataset is available in the openneuro repository at https://openneuro.org/datasets/ds000031/versions/2.0.2.
Data from the Midnight Scan Club dataset is available in the openneuro repository at https://openneuro.org/datasets/ds000224/versions/1.0.4. Data from the Cast-
induced plasticity dataset is available in the openneuro repository at https://openneuro.org/datasets/ds002766/versions/3.0.0.
Data from the Natural Scenes Dataset is available from Amazon Web Services (AWS) at https://registry.opendata.aws/nsd/.
The ABCD data used in this report are from Annual Release 5.0.
Data from individual subjects with depression are a part of ongoing clinical trials and not publicly available at this time.
Findings apply to all studied individuals and groups, regardless of sex.
Sex ratios:
Serial imaging of Major Depression Dataset: 3M, 3F
Weill Cornell Medicine rTMS 1 dataset: 21M, 27F
Weill Cornell Medicine rTMS 2 dataset: 24M, 21F
Stanford University rTMS dataset: 13M, 29 F
Weill Cornell Late-onset Depression dataset: 0M, 5F
Weill Cornell Multi-echo dataset: 7M, 0F
MyConnectome dataset: 1M, 0F
Midnight Scan Club: 5M, 5F
Cast-induced plasticity dataset: 1M, 0F
Natural Scenes Dataset: 2M, 6F
Eskalibur dataset: 5M, 5F
Sample selected from Adolescent Brain Cognitive Development study: 54M, 60F
No socially constructed or socially relevant categorization variables were used or are relevant for our manuscript.
Findings apply to all studied individuals and groups, regardless of age.
Age:
Serial imaging of Major Depression Dataset: mean age = 29.47 ± 8.28 years
Weill Cornell Medicine rTMS 1 dataset: mean age = 40.89 ± 12.73 years
Weill Cornell Medicine rTMS 2 dataset: mean age = 44.46 ± 15.38 years
Stanford University rTMS dataset: mean age = 38.09 ± 12.77 years
Weill Cornell Late-onset Depression dataset: mean age = 66.60 ± 5.31 years
Weill Cornell Multi-echo dataset: mean age = 33.42 ± 9 years
MyConnectome dataset: 45 years-old
Midnight Scan Club: mean age = 29.1 ± 3.3 years
Cast-induced plasticity dataset: 27 years-old
Natural Scenes Dataset: mean age = 26.50 ± 4.24 years
Eskalibur dataset: mean age = 31.4 ± 5.4 years
Adolescent Brain Cognitive Development dataset: 9.46 ± 0.50 years at baseline study visit
Serial imaging of Major Depression Dataset: Individuals with depression were recruited from the NYC metro area via flyers
and word of mouth.
Weill Cornell rTMS dataset: Individuals with depression were recruited from the NYC metro area via flyers and word of
mouth.
Stanford University rTMS dataset: Individuals with depression were recruited from the Bay area via flyers and word of mouth.
Content courtesy of Springer Nature, terms of use apply. Rights reserved

3
nature
portfolio
|
reporting
summary
April
2023
Ethics oversight
Note that full information on the approval of the study protocol must also be provided in the manuscript.
Field-specific reporting
Please select the one below that is the best fit for your research. If you are not sure, read the appropriate sections before making your selection.
Life sciences Behavioural & social sciences Ecological, evolutionary & environmental sciences
For a reference copy of the document with all sections, see nature.com/documents/nr-reporting-summary-flat.pd f
Life sciences study design
All studies must disclose on these points even when the disclosure is negative.
Sample size
Data exclusions
Replication
Randomization
Blinding
Reporting for specific materials, systems and methods
We require information from authors about some types of materials, experimental systems and methods used in many studies. Here, indicate whether each material,
system or method listed is relevant to your study. If you are not sure if a list item applies to your research, read the appropriate section before selecting a response.
Weill Cornell Late-onset Depression dataset: Individuals with depression were recruited from the Bay area via flyers and word
of mouth.
Weill Cornell Multi-echo dataset: Healthy adult subjects were recruited from the Weill Cornell Medical School community via
word of mouth.
MyConnectome dataset: The subject in this dataset was also the principal investigator.
Midnight Scan Club: Healthy adult subjects were recruited from the Washington University community via flyers and word of
mouth.
Cast-induced plasticity dataset: Healthy adult subjects were recruited from the Washington University community via flyers
and word of mouth.
Natural Scenes Dataset: Participants were recruited through advertisements to the local community and were screened
based on ability to participate in a neuroimaging study.
Eskalibur dataset: Participants were recruited through advertisements to the local community.
Adolescent Brain Cognitive Development dataset: Participants were recruited from a nationally distributed set of 21 study
sites.
Serial imaging of Major Depression Dataset, Weill Cornell rTMS dataset, Stanford University rTMS dataset, Weill Cornell Late-
onset Depression dataset, and Weill Cornell Multi-echo dataset: The study was approved by the Weill Cornell Medicine
Institutional Review Board.
MyConnectome dataset: It was determined that this study did not meet requirements for human subjects research and thus
that Institutional Review Board (IRB) approval was not necessary. Subsequent data collection at Washington University was
collected under an approved IRB protocol.
Midnight Scan Club, Cast-induced plasticity dataset: These studies were approved by the Washington University School of
Medicine Human Studies Committee and Institutional Review Board.
Natural Scenes Dataset: Participants were recruited through advertisements to the local community and were screened
based on ability to participate in a neuroimaging study.
Eskalibur dataset: The study was approved by the local ethics committee.
Adolescent Brain Cognitive Development dataset: The ABCD Study obtained centralized institutional review board approval
from the university of California, San Diego, and each of the 21 study sites obtained local institutional review board approval.
This study collected large quantities of data (repeated fMRI and clinical assessments) in individual subjects, and the majority of our analyses
are conducted at the within-subject level. For these analyses, the relevant factor is having enough high quality data per-subject for reliable
and accurate inferences. We and other groups (i.e., Laumann et al., 2015, Gordon et al., 2017, Lynch et al., 2020) have shown that
approximately 30 minutes of resting-state fMRI data per-subject is necessary for reliable functional connectivity measurements, and we have
at least that much data for each subject in our study.
No subjects were excluded from analyses.
The main experimental findings (Figure 1c) were replicated (thrice) in held-out replication datasets, and in independent large n group-average
datasets (Extended Data Figure 4). Brain-behavior relationships observed in individual subjects studied longitudinally (e.g., Figure 3, Figure 4)
were replicated across multiple subjects.
Experimental groups consisted of whether or not individual subjects had a diagnosis of major depression. Therefore, randomization is not
possible.
No blinding was performed.
Content courtesy of Springer Nature, terms of use apply. Rights reserved

4
nature
portfolio
|
reporting
summary
April
2023
Materials & experimental systems
n/a Involved in the study
Antibodies
Eukaryotic cell lines
Palaeontology and archaeology
Animals and other organisms
Clinical data
Dual use research of concern
Plants
Methods
n/a Involved in the study
ChIP-seq
Flow cytometry
MRI-based neuroimaging
Clinical data
Policy information about clinical studies
All manuscripts should comply with the ICMJEguidelines for publication of clinical research and a completedCONSORT checklist must be included with all submissions.
Clinical trial registration
Study protocol
Data collection
Outcomes
Magnetic resonance imaging
Experimental design
Design type
Design specifications
Behavioral performance measures
Acquisition
Imaging type(s)
Field strength
Sequence & imaging parameters
NCT04041479; NCT04982757
Study overview is available online (NCT04041479: https://clinicaltrials.gov/study/NCT04041479#study-overview; NCT04982757:
https://classic.clinicaltrials.gov/ct2/show/NCT04982757
NCT04041479: Data was collected at Weill Cornell Medicine and Stanford University starting on 9/17/2021. NCT04982757: Data was
collected at Weill Cornell Medicine starting on 7/29/2021. Data collection is ongoing as of 5/30/2024.
NCT04041479: Primary outcome is change in depression, as measured by the Hamilton Depression Rating Scale (HAMD17).
Secondary outcome is change in depression, as measured by the Quick Inventory of Depressive Symptomatology (QIDS). The trial is
ongoing, the trial's primary outcome measure was not analyzed in the present study.
NCT04982757: Primary outcome is percent change in Montgomery-Asberg Depression Rating Scale (MADRS) scores for participants
with treatment resistant depression [ Time Frame: Baseline to Treatment End: Day 5 or 10 (depending on number of 5-day treatment
courses administered) ]. The MADRS is a measure of depression symptoms and is scored on a scale of 0 to 60, with 0 being no
depressive symptoms and 60 being severe depressive symptoms. The secondary outcome measure is Percent Change in Quick
Inventory of Depressive Symptomatology (QIDS) scores for participants with OCD [ Time Frame: Baseline to Treatment End: Day 5 or
10 (depending on number of 5-day treatment courses administered) ]. The QIDS is a self-report measure of depression symptoms
and is scored on a scale of 0 to 27, with 0 being no depressive symptoms and 27 being severe depressive symptoms. The trial is
ongoing, the trial's primary outcome measure was not analyzed in the present study.
Resting-state fMRI
Resting-state fMRI: 47 to 1,792 minutes of data per-subject
Behavioral outputs were not recorded.
Structural, Functional.
3T, 7T
Serial Imaging of Major Depression dataset: MRI data were acquired on a Siemens Magnetom Prisma 3T scanner at the
Citigroup Biomedical Imaging Center of Weill Cornell's medical campus using a Siemens 32-channel head coil. Multi-
echo, multi-band resting-state fMRI scans were collected using a T2*-weighted echo-planar sequence covering the full
brain (TR: 1355 ms; TE1: 13.40 ms, TE2: 31.11 ms, TE3: 48.82 ms, TE4: 66.53 ms, and TE5: 84.24 ms; FOV: 216 mm; flip
angle: 68° (the Ernst angle for gray matter assuming a T1 value of 1400 ms); 2.4 mm isotropic voxels; 72 slices; AP phase
encoding direction; in-plane acceleration factor: 2; and multi-band acceleration factor: 6) with 640 volumes acquired
per scan for a total acquisition time of 14 minutes and 27 seconds. Spin echo EPI images with opposite phase encoding
directions (AP and PA) but identical geometrical parameters and echo spacing were acquired before each resting-state
scan. Multi-echo T1-weighted (TR/TI: 2500/1000 ms; TE1: 1.7 ms, TE2: 3.6 ms, TE3: 5.5 ms, TE4: 7.4 ms ; FOV: 256 mm;
flip angle: 8°, and 208 sagittal slices with a 0.8 mm slice thickness) and T2-weighted anatomical images (TR: 3200 ms; TE:
563 ms; FOV: 256; flip angle: 8°, and 208 sagittal slices with a 0.8 mm slice thickness) were also collected.
Weill Cornell rTMS dataset: MRI data were acquired on a Siemens Magnetom Prisma 3T machine at the Citigroup
Biomedical Imaging Center of Weill Cornell's medical campus using a Siemens 32-channel head coil. Multi-echo, multi-
Content courtesy of Springer Nature, terms of use apply. Rights reserved

5
nature
portfolio
|
reporting
summary
April
2023
Area of acquisition
Diffusion MRI Used Not used
Preprocessing
Preprocessing software
Normalization
Normalization template
Noise and artifact removal
band
resting-state
fMRI
scans
were
collected
at
each
study
visit
using
a
T2*-weighted
echo-planar
sequence
covering
the full brain (TR: 1300 ms; TE1: 12.60 ms, TE2: 29.51 ms, TE3: 46.42 ms, and TE4: 63.33 ms; FOV: 216 mm; flip angle:
67° (the Ernst angle for gray matter assuming a T1 value of 1400 ms); 2.5 mm isotropic voxels; 60 slices; AP phase
encoding direction; in-plane acceleration factor: 2; and multi-band acceleration factor: 4) with 650 volumes acquired
per scan for a total acquisition time of 14 minutes and 5 seconds. Spin echo EPI images with opposite phase encoding
directions (AP and PA) but identical geometrical parameters and echo spacing were acquired before each resting-state
scan. Multi-echo T1-weighted (TR/TI: 2500/1000 ms; TE1: 1.7 ms, TE2: 3.6 ms, TE3: 5.5 ms, TE4: 7.4 ms ; FOV: 256; flip
angle: 8°, and 208 sagittal slices with a 0.8 mm slice thickness) and T2-weighted anatomical images (TR: 3200 ms; TE:
563 ms; FOV: 256; flip angle: 8°, and 208 sagittal slices with a 0.8 mm slice thickness) were also collected.
Stanford University rTMS dataset: MRI data were acquired on a GE SIGNA 3T machine at the Center for Neurobiological
Imaging on Stanford University’s campus using a Nova Medical 32-channel head coil. Multi-echo, multi-band resting-
state fMRI scans were collected using a T2*-weighted echo-planar sequence covering the full brain (TR: 1330 ms; TE1:
13.7 ms, TE2: 31.60 ms, TE3: 49.50 ms, and TE4: 67.40 ms; flip angle: 67° (the Ernst angle for gray matter assuming a T1
value of 1400 ms); 3 mm isotropic voxels; 52 slices; AP phase encoding direction; in-plane acceleration factor: 2; and
multi-band acceleration factor: 4) with 338 volumes acquired per scan for a total acquisition time of 7 minutes and 30
seconds. Spin echo EPI images with opposite phase encoding directions (AP and PA) but identical geometrical
parameters and echo spacing were acquired before each resting-state scan.T1-weighted and T2-weighted anatomical
images were also collected.
Weill Cornell Late-onset Depression dataset: MRI data were acquired on a Siemens Magnetom Prisma 3T machine at the
Citigroup Biomedical Imaging Center of Weill Cornell's medical campus using a Siemens 32-channel head coil. Multi-
echo, multi-band resting-state fMRI scans were collected at each study visit using a T2*-weighted echo-planar sequence
covering the full brain (TR: 1300 ms; TE1: 12.60 ms, TE2: 29.51 ms, TE3: 46.42 ms, and TE4: 63.33 ms; FOV: 216 mm;
flip angle: 67° (the Ernst angle for gray matter assuming a T1 value of 1400 ms); 2.5 mm isotropic voxels; 60 slices; AP
phase encoding direction; in-plane acceleration factor: 2; and multi-band acceleration factor: 4) with 480 volumes
acquired per scan for a total acquisition time of 10 minutes and 38 seconds. Spin echo EPI images with opposite phase
encoding directions (AP and PA) but identical geometrical parameters and echo spacing were acquired before each
resting-state scan. Multi-echo T1-weighted (TR/TI: 2500/1000 ms; TE1: 1.7 ms, TE2: 3.6 ms, TE3: 5.5 ms, TE4: 7.4 ms ;
FOV: 256 mm; flip angle: 8°, and 208 sagittal slices with a 0.8 mm slice thickness) and T2-weighted anatomical images
(TR: 3200 ms; TE: 563 ms; FOV: 256; flip angle: 8°, and 208 sagittal slices with a 0.8 mm slice thickness) were also
acquired.
Whole-brain
Preprocessing of multi-echo data minimized spatial interpolation and volumetric smoothing while preserving the alignment of
echoes. The single-band reference (SBR) images (one per echo) for each scan were averaged. The resultant average SBR
images were aligned, averaged, co-registered to the ACPC aligned T1-weighted anatomical image, and simultaneously
corrected for spatial distortions using FSL’s topup and epi_reg programs. Freesurfer’s bbregister algorithm was used to refine
this co-registration. For each scan, echoes were combined at each timepoint and a unique 6 DOF registration (one per
volume) to the average SBR image was estimated using FSL’s MCFLIRT tool, using a 4-stage (sinc) optimization. All of these
steps (co-registration to the average SBR image, ACPC alignment, and correcting for spatial distortions) were concatenated
using FSL’s convertwarp tool and applied as a single spline warp to individual volumes of each echo after correcting for slice
time differences using FSL’s slicetimer program. The functional images underwent a brain extraction using the co-registered
brain extracted T1-weighted anatomical image as a mask and corrected for signal intensity inhomogeneities using ANT’s
N4BiasFieldCorrection tool.
Software packages incorporated into the preprocessing pipelines included: Matlab R2019a, https://www.mathworks.com/;
Connectome Workbench 1.4.2, http://www.humanconnectome.org/software/connectome-workbench.html; Freesurfer v6,
https://surfer.nmr.mgh.harvard.edu/; FSL 6.0, https://fsl.fmrib.ox.ac.uk/fsl/fslwiki; and Infomap, https://
www.mapequation.org.
T1w ---> atlas linear, BOLD --> atlas.
MNI
Preprocessed multi-echo data were submitted to multi-echo ICA (ME-ICA), which is designed to isolate spatially structured
T2*- (neurobiological; “BOLD-like”) and S0-dependent (non-neurobiological; “not BOLD-like”) signals and implemented using
the “tedana.py” workflow 65. In short, the preprocessed, ACPC-aligned echoes were first combined according to the average
rate of T2* decay at each voxel across all time points by fitting the monoexponential decay, S(t) = S0e -t / T2* . From these
T2* values, an optimally-combined multi-echo (OC-ME) time-series was obtained by combining echoes using a weighted
average (WTE = TE * e -TE/ T2*). The covariance structure of all voxel time-courses was used to identify major signals in the
OC-ME time-series using principal component and independent component analysis. Components were classified as either
T2*-dependent (and retained) or S0-dependent (and discarded), primarily according to their decay properties across echoes.
All component classifications were manually reviewed by author CJL and revised when necessary. Mean gray matter time-
series regression was performed to remove spatially diffuse noise. Temporal masks were generated for censoring high
motion time-points using a framewise displacement (FD) threshold of 0.3 mm and a backward difference of two TRs, for an
effective sampling rate comparable to historical FD measurements (approximately 2 to 4 seconds). Prior to the FD calculation,
head realignment parameters were filtered using a stopband Butterworth filter (0.2 - 0.35 Hz) to attenuate the influence of
Content courtesy of Springer Nature, terms of use apply. Rights reserved

6
nature
portfolio
|
reporting
summary
April
2023
Volume censoring
Statistical modeling & inference
Model type and settings
Effect(s) tested
Specify type of analysis: Whole brain ROI-based Both
Anatomical location(s)
Statistic type for inference
(See Eklund et al. 2016)
Correction
Models & analysis
n/a Involved in the study
Functional and/or effective connectivity
Graph analysis
Multivariate modeling or predictive analysis
Functional and/or effective connectivity
Graph analysis
Multivariate modeling and predictive analysis
respiration
on
motion
parameters.
Single-echo
fMRI
datasets
(e.g.,
NSD,
ABCD)
were
subjected
to
the
same
preprocessing
procedures, except ICA-AROMA was used instead of ME-ICA.
See above.
We measured the surface area (in mm2) that each vertex in the individual’s midthickness surface is responsible for
(“wb_command --surface-vertex-areas”). Next, we calculated the relative contribution (size) of each functional network to
the total cortical surface area by taking the total surface area of all network vertices in relation to the total cortical surface
area. In the striatum, where each voxel represents the same amount of tissue, the relative contribution of each functional
network to the total striatal volume was calculated by taking the total number of network voxels in relation to the total
striatal voxels.The statistical significance of group differences in network size were evaluated using permutation tests and
independent sample t-tests (the latter implemented using Matlab’s ttest2.m function). Effect size (Cohen’s d) was calculated
as difference in group means divided by pooled standard deviation. Assumptions regarding equal variance were adjusted
when appropriate (based on two-sample F-tests performed using Matlab’s vartest2.m function). The relative difference
between groups was calculated as the absolute difference divided by network size in healthy controls.
With respect to testing differences in functional network size, permutation and independent sample t-tests tested against
null hypothesis that difference in group means is zero.
Individual-specific functional networks were created from each individual's resting-state fMRI data.
No cluster wise inferences were made.
P-values were adjusted using Bonferroni correction for multiple comparisons.
Pearson correlation.
Individual-specific (and in some cases, group-average) functional networks were identified using the
procedures described in Gordon et al., 2020 PNAS and Lynch et al., 2022 Neuron. A functional connectivity
matrix summarizing the correlation between the time-courses of all cortical vertices and subcortical voxels
across all study visits was constructed. Correlations between nodes 10 mm apart (geodesic and Euclidean
space used for cortico-cortical and subcortical-cortical distance, respectively) were set to zero. Correlations
between voxels belonging to subcortical structures were set to zero. Functional connectivity matrices were
thresholded in such a way that they retained at least the strongest X% correlations (0.01, 0.02, 0.05, 0.1, 0.2,
0.5, 1, 2, and 5%) to each vertex and voxel and were used as inputs for the InfoMap community detection
algorithm. Free parameters (for example, the number of algorithm repetitions) for the Infomap algorithm
were fixed across subjects. The optimal scale for further analysis across individuals was defined as the graph
threshold producing the best size-weighted average homogeneity relative to the median of the size-weighted
average homogeneity calculated from randomly rotated networks. Size-weighted average homogeneity was
maximized relative to randomly rotated communities at the 0.1% graph density. Each Infomap community
was algorithmically assigned to one of 20 possible functional network identities (Default-Parietal, Default-
Anterolateral, Default-Dorsolateral, Default-Retrosplenial, Visual-Lateral, Visual-Stream, Visual-V1, Visual-V5,
Frontoparietal, Dorsal Attention, Premotor / Dorsal Attention II, Language, Salience, Cingulo-opercular /
Action-mode, Parietal memory, Auditory, Somatomotor-Hand, Somatomotor-Face, Somatomotor-Foot,
Auditory, or Somato-Cognitive-Action) primarily according their functional connectivity and spatial locations
relative to a specified set of priors. This procedure is implemented using a Matlab function
(“pfm_identify_networks.m”) available on our study’s GitHub repository (https://github.com/cjl2007/PFM-
Depression). The priors used in our study (“Priors.mat”) are also available in the same repository. All
algorithmic assignments were manually reviewed by study author CJL and manually adjusted in the case of
an ambiguous assignment.
A support vector machine classifier, where class labels were diagnosis status (healthy control or depression)
and features were functional brain network size in each individual, was trained using repeated (100
iterations) nested split-half (2-fold) cross-validation with a grid search optimization strategy for
hyperparameter tuning (box constraint, kernel size). Classification accuracy was calculated as the percentage
of correct predictions, and statistical significance assessed using permutation tests. Confusion matrix was
created using Matlab’s confmat.m function. Feature importance was evaluated by iteratively omitting each
functional network and calculating the resulting loss in accuracy. The Synthetic Minority Oversampling
Technique (SMOTE) was used to prevent classification bias in favor of the majority class, and performed on
Content courtesy of Springer Nature, terms of use apply. Rights reserved

7
nature
portfolio
|
reporting
summary
April
2023
training
data
only
to
to
avoid
data
leakage.
Content courtesy of Springer Nature, terms of use apply. Rights reserved

1.
2.
3.
4.
5.
6.
Terms and Conditions
Springer Nature journal content, brought to you courtesy of Springer Nature Customer Service Center GmbH (“Springer Nature”).
Springer Nature supports a reasonable amount of sharing of research papers by authors, subscribers and authorised users (“Users”), for small-
scale personal, non-commercial use provided that all copyright, trade and service marks and other proprietary notices are maintained. By
accessing, sharing, receiving or otherwise using the Springer Nature journal content you agree to these terms of use (“Terms”). For these
purposes, Springer Nature considers academic use (by researchers and students) to be non-commercial.
These Terms are supplementary and will apply in addition to any applicable website terms and conditions, a relevant site licence or a personal
subscription. These Terms will prevail over any conflict or ambiguity with regards to the relevant terms, a site licence or a personal subscription
(to the extent of the conflict or ambiguity only). For Creative Commons-licensed articles, the terms of the Creative Commons license used will
apply.
We collect and use personal data to provide access to the Springer Nature journal content. We may also use these personal data internally within
ResearchGate and Springer Nature and as agreed share it, in an anonymised way, for purposes of tracking, analysis and reporting. We will not
otherwise disclose your personal data outside the ResearchGate or the Springer Nature group of companies unless we have your permission as
detailed in the Privacy Policy.
While Users may use the Springer Nature journal content for small scale, personal non-commercial use, it is important to note that Users may
not:
use such content for the purpose of providing other users with access on a regular or large scale basis or as a means to circumvent access
control;
use such content where to do so would be considered a criminal or statutory offence in any jurisdiction, or gives rise to civil liability, or is
otherwise unlawful;
falsely or misleadingly imply or suggest endorsement, approval , sponsorship, or association unless explicitly agreed to by Springer Nature in
writing;
use bots or other automated methods to access the content or redirect messages
override any security feature or exclusionary protocol; or
share the content in order to create substitute for Springer Nature products or services or a systematic database of Springer Nature journal
content.
In line with the restriction against commercial use, Springer Nature does not permit the creation of a product or service that creates revenue,
royalties, rent or income from our content or its inclusion as part of a paid for service or for other commercial gain. Springer Nature journal
content cannot be used for inter-library loans and librarians may not upload Springer Nature journal content on a large scale into their, or any
other, institutional repository.
These terms of use are reviewed regularly and may be amended at any time. Springer Nature is not obligated to publish any information or
content on this website and may remove it or features or functionality at our sole discretion, at any time with or without notice. Springer Nature
may revoke this licence to you at any time and remove access to any copies of the Springer Nature journal content which have been saved.
To the fullest extent permitted by law, Springer Nature makes no warranties, representations or guarantees to Users, either express or implied
with respect to the Springer nature journal content and all parties disclaim and waive any implied warranties or warranties imposed by law,
including merchantability or fitness for any particular purpose.
Please note that these rights do not automatically extend to content, data or other material published by Springer Nature that may be licensed
from third parties.
If you would like to use or distribute our Springer Nature journal content to a wider audience or on a regular basis or in any other manner not
expressly permitted by these Terms, please contact Springer Nature at
onlineservice@springernature.com