Available via license: CC BY 4.0
Content may be subject to copyright.
RESEARCH ARTICLE
Machine-learning a virus assembly fitness
landscape
Pierre-Philippe DechantID
1,2,3☯
*, Yang-Hui He
4,5,6☯
1School of Science, Technology & Health, York St John University, York, United Kingdom, 2York Cross-
disciplinary Centre for Systems Analysis, University of York, Heslington, United Kingdom, 3Department of
Mathematics, University of York, Heslington, United Kingdom, 4Department of Mathematics, City, University
of London, London, United Kingdom, 5Merton College, University of Oxford, Oxford, United Kingdom,
6School of Physics, NanKai University, Tianjin, P.R. China
☯These authors contributed equally to this work.
*ppd22@cantab.net
Abstract
Realistic evolutionary fitness landscapes are notoriously difficult to construct. A recent cut-
ting-edge model of virus assembly consists of a dodecahedral capsid with 12 corresponding
packaging signals in three affinity bands. This whole genome/phenotype space consisting
of 3
12
genomes has been explored via computationally expensive stochastic assembly mod-
els, giving a fitness landscape in terms of the assembly efficiency. Using latest machine-
learning techniques by establishing a neural network, we show that the intensive computa-
tion can be short-circuited in a matter of minutes to astounding accuracy.
1 Introduction
Two facts about simple viruses have been known for a long time. Firstly, that genetic economy
leads to the use of symmetry, such that virus capsids are mostly icosahedral or helical. Sec-
ondly, packaging signals, that is secondary structure features in the viral RNA, are often
required for encapsidation in viruses with single-stranded genomes. Examples are the Origin
of Assembly (OAS) sequence in Tobacco Mosaic Virus (TMV), the psi element in HIV
(Human Immunodeficiency Virus) and the TR sequence (Translational Repressor) in MS2
(Male Specific 2 bacteriophage). This is an evolutionary advantage, as it ensures vRNA-specific
encapsidation and can increase assembly efficiency through a cooperative role of the RNA,
which acts as a nucleation site.
More recently, it has been shown that taken together, these two facts suggest that there
could be more than one packaging signal, with multiple signals in fact dispersed throughout
the genome [1,2]. This is because the capsid is symmetric, and the packaging signal mecha-
nism functions via interaction between viral RNA and the coat protein (CP). In several cases,
this RNA-CP interaction leads to a conformational change in the CP, which only then makes
it assembly competent (e.g. TMV and MS2 [3]. The picture that emerges is then that there are
multiple packaging signals (PS) that recruit CP onto a growing capsid. This reduces the phase
space that CP has to search in order to assemble a capsid, resulting in vastly increased assembly
efficiency. The details of such a mechanism were found in MS2 and STNV (Satellite Tobacco
PLOS ONE
PLOS ONE | https://doi.org/10.1371/journal.pone.0250227 May 5, 2021 1 / 10
a1111111111
a1111111111
a1111111111
a1111111111
a1111111111
OPEN ACCESS
Citation: Dechant P-P, He Y-H (2021) Machine-
learning a virus assembly fitness landscape. PLoS
ONE 16(5): e0250227. https://doi.org/10.1371/
journal.pone.0250227
Editor: Chi-Hua Chen, Fuzhou University, CHINA
Received: July 13, 2019
Accepted: April 1, 2021
Published: May 5, 2021
Copyright: ©2021 Dechant, He. This is an open
access article distributed under the terms of the
Creative Commons Attribution License, which
permits unrestricted use, distribution, and
reproduction in any medium, provided the original
author and source are credited.
Data Availability Statement: The fitness landscape
data has been provided by Twarock, Dykeman and
Bingham. All our analysis is reproducible, but the
landscape has not been published and is owned by
Twarock, Dykeman, and Bingham. The data was
obtained through personal communication so
others could politely request the same information
through the authors e.g. Richard Bingham (r.j.
bingham@york.ac.uk), Eric Dykeman (eric.
dykeman@york.ac.uk) or Reidun Twarock (reidun.
twarock@york.ac.uk). The authors did not have any
special access privileges that others would not
have.
Necrosis Virus) as model systems. Once the details of this mechanism were understood using
biochemistry, structural biology [4], bioinformatics [5], biophysics and graph theory [6] in
these model systems, related mechanisms could be found in clinically relevant viruses such as
Hepatitis C virus (HCV), Hepatitis B virus (HBV) and Human Parechovirus. These packaging
signals are secondary structure features of the viral genomes where a stemloop in the single-
stranded RNA presents a common recognition motif that can bind to CP (see Fig 1A). The
viral genome thus has multiple layers of constraints, by having to code for genes as well as the
PS instruction manual. This set of packaging signals can also be repurposed and optimised for
the assembly of virus-like particles, which do not share the same genetic constraints as the
virus, and could be used e.g. for vaccines, drug delivery or as an anti-viral strategy [7].
An equilibrium model of how to assemble a simplified example of an icosahedral virus, a
dodecahedron built from 12 pentagonal faces, was considered in [8] using an ODE (ordinary
differential equations) model. More recently, the multiple dispersed packaging signal para-
digm has sparked renewed interested in such a dodecahedral model [9]. The assembly reaction
kinetics was modelled via a set of discrete reactions in a stochastic simulation paradigm based
on the Gillespie algorithm [10].
In this model 12 PSs can bind CP, as well as dissociate again, reflecting reversible/equilib-
rium kinetics. Bound CP can then bind other bound CP, gradually building up a capsid (see
Fig 1Band 1C). The PSs here have three different bands of binding affinity: weak, medium
and strong. These choices correspond to binding energies of 4/8/12 kcal/mol respectively,
Fig 1. A The nucleotide sequence of a virus determines the gene products; however, in addition to this information content the RNA also explores a configuration
space of secondary structures. Viruses appear to have evolved to use such motifs to help recruit coat protein with a conserved common recognitionmotif during
assembly. The stability and binding affinity of these packaging signals gives a distinctive profile for viral assembly, which is the phenotype relevant to assembly.
Assembly efficiency is the fitness of this phenotype, or at least the contribution to the overall fitness that is determined by aspects of assembly. BThe genomes in the
model consist of twelve packaging signals (PS) that can take weak, medium and strong binding affinities. They successively recruit twelve pentagonal coat proteins,
which together form the dodecahedral virion in this model. CThe stochastic simulation algorithm models several possible reactions. Firstly, packaging signals can bind
coat proteins (and fall apart again), and secondly, two coat proteins that have been recruited by packaging signals can bind to each other. The fitness landscape was
computed for 2000 virions for each possible genome, making the computations very intensive.
https://doi.org/10.1371/journal.pone.0250227.g001
PLOS ONE
Machine-learning a virus assembly fitness landscape
PLOS ONE | https://doi.org/10.1371/journal.pone.0250227 May 5, 2021 2 / 10
Funding: This study was supported by the Science
and Technology Facilities Council (STFC) (https://
stfc.ukri.org/) in the form of a grant awarded to
YHH (ST/J00037X/1).
Competing interests: The authors have declared
that no competing interests exist.
based on the TR sequence in MS2 which has approximately 12kcal/mol. The binding energy
between CPs is much lower, at approximately 2 kcal/mol. This modulation of affinity affects
the assembly kinetics, e.g. by providing a nucleation point that starts assembly, or allowing for
error-correction via weaker binding elsewhere. The thermodynamics of PS binding and of the
number of CP bonds formed then translates into assembly efficiency. This in turn is taken as a
proxy for fitness (all other things being equal)—or at least the contribution to the fitness that
results from assembly considerations [11–13].
In [14] the whole space of these 3
12
genomes (or rather, phenotype profiles) has been
explored. The assembly efficiency there is given by the number of capsids that have correctly
assembled out of a possible total of 2, 000. This efficiency provides a fitness landscape on the
12-dimensional genome space. This is an interesting model that is tractable, in contrast with
many other biological systems, as it has a small number of degrees of freedom and is domi-
nated by the symmetry of the capsid. This tractability also allows for the consideration of viral
evolution. For instance, mutation of the PS strengths leads to the emergence of a set of related
genomes that form a ‘quasispecies’ [15,16]. One can thus investigate the effect of evolutionary
pressures, e.g. those exerted by standard drugs or a novel type of drug that targets packaging
signals [9].
This model thus captures many interesting aspects of viral genetics, geometry and assembly.
A more realistic model would have more CP building blocks and PSs, e.g. around 60 for MS2
(i.e. one full orbit of the icosahedral group). But the computation time for even these simple
genomes and the assembly kinetics that provide the fitness landscape are already considerable.
Even other simplified models, e.g. reduced orbits on symmetry axes given by e.g. an icosahe-
dron consisting of 20 triangles with 20 PSs, a rhombic triacontahedron consisting of 30 rhom-
buses with 30 PSs, or a finer gradation of binding affinity bands are already computationally
out of reach. For experimental approaches to measure local fitness values please see for
instance [17], though note that in this reference this is limited to only 48 data points, whereas
the fitness space of the simple above model is already 10; 000 times that.
2 Results/Discussion
This data set is a perfect example of data that is amenable to a machine-learning approach,
since it associates a vector input with a number output. We therefore train a neural network to
predict the fitness landscape. The network is trained on a subset of the whole genome space,
and validated on the remainder of the data. This proof-of-principle shows that it is very fast for
a neural network to learn the inherent patterns within the large degeneracy of the detailed sto-
chastic modelling to predict assembly efficiency fitness for unseen genomes to extremely high
accuracy (c.f. the paper [18] which has been published after submission of this manuscript that
applies machine learning to the related problem of finding high-level behaviour within the
large degeneracies of protein folding). The danger is that some subtleties of the stochastic
modelling are lost, but allowing for computation times many orders of magnitude faster. More
likely, however, since the data is obtained from a Monte Carlo simulation, and many works in
the literature immediately go to an ODE approximation and miss these details anyway, the ML
is actually indifferent to this. In fact, the discontinuities from the stochastic method, which the
ODE smoothing would not capture, is perfectly adapted to the neural network and machine
learning classifiers, which are much better adapted to partitioning fitness space in more subtle
ways. Many different neural network architectures were tested and all led to very similar
results. This supports the point that there is something reasonably simple underlying the data
set that can be learned by any reasonable neural network. This approach could thus in future
be used to tackle more realistic models such as the ones mentioned above. Stochastic
PLOS ONE
Machine-learning a virus assembly fitness landscape
PLOS ONE | https://doi.org/10.1371/journal.pone.0250227 May 5, 2021 3 / 10
simulations could be used to partially explore these larger genome spaces, calculating assembly
fitness in order to provide a training set for a neural network. The rest of the fitness landscape
can then be predicted by the artificial intelligence; it is also possible to only compute this fit-
ness if necessary, e.g. when a new genome arises through mutation in a quasispecies model,
such that such computation may only be necessary ‘precedurally’. By that we mean that they
are only calculated locally when required during the computation, e.g. when evolutionary
dynamics starts exploring a certain range in fitness space, and not calculating the entire fitness
space before beginning the simulation.
The assembly process discussed here is ultimately a problem of geometry and thermody-
namics. So it would with some modification also apply to the assembly of other icosahedral
particles such as virus-like particles (VLPs) for drug delivery or as vaccines, as well as to carbon
fullerene assembly, which are very attractive fields for biomedical and nanoscience applica-
tions. For instance, virus-like particles could present viral epitopes in order to act as vaccines.
In order to find the most efficient assembly pathway for such VLPs, an analogue of the above
fitness landscape could be constructed in order to solve the resulting optimisation problem
and to give industry suggestions which parts of the fitness landscape to explore deeper experi-
mentally.
3 Methods
From a purely mathematical point of view, we have the following problem. Let (weak,
medium, strong) be denoted respectively by (1, 2, 3). The input is a vector vin a 12-dimen-
sional vector space over
3
, the field of three elements. The output is an integer (which we treat
as a real number) between 0 and 2000, which we can normalise into �2[0, 1] by dividing by
2000. The algorithm used by [14] is thus a map
f:v212
3 !�2 ½0;1:ð1Þ
A typical example is
f1;1;1;2;2;2;3;1;2;2;1;1g !1523
2000 ’0:7615 ð2Þ
3.1 Computational aspects of the simulation
The map fis a computationally intensive one with individual genome run times between 20
minutes and 12 hours, and cumulative run time of 3-4 weeks on the N8 Polaris high perfor-
mance computing research cluster, Intel 2.6 GHz Sandy Bridge E5-2670 processors, with a
total of 5, 312 cores, with a mix of 4 and 16Gb of RAM (https://n8hpc.org.uk/facilities/) [9].
The fluctuations in numbers of assembled capsids tend to be in the tenth of a percent range
(i.e. ±20 virions), however when initially running the code with 75 repeats of each run for cer-
tain genomes the standard error was very small (±0.001%) [9]. Due the amount of time and
cluster resources it already took, each point on the landscape is the result of only 2 simulation
runs. While not ideal, cluster computation was limited; the simulation allows to capture the
more generic features whilst the above cross-validation suggests that the error is small [9].
Nevertheless a brute-force simulation has been performed on the 3
12
= 531, 441 possible input
values and the efficiency value extracted. This gives us a database of some half a million known
cases of the form (2).
PLOS ONE
Machine-learning a virus assembly fitness landscape
PLOS ONE | https://doi.org/10.1371/journal.pone.0250227 May 5, 2021 4 / 10
3.2 Machine-learning the dataset
Such a problem is perfectly adapted to supervised machine-learning: we know many input val-
ues and wish to train some artificial intelligence to associate the input with the known output
on some small subset, and use it to predict the output for unseen input [19]. The advantage of
this approach is that often approximate results can be attained at reduction in computation
time by many orders of magnitude. The paradigm of using machine-learning in algebraic
geometry and more general classes of problems in pure mathematics was proposed in [20–22]
to satisfying accuracy, and it is a similar philosophy that we will adopt here.
Let us first try the following specific procedure:
• Take the full data Dof the form (2), of size 3
12
;
• Establish the neural network, a 3-layer perceptron
INPUT ¼v!L20 !S20 !L20 !S1!OUTOUT ¼�
INPUT ¼v!L20 !S20 !L20 !S1!OUTOUT ¼�
• In the above, L means a linear-layer, S, a sigmoid layer and S, a summation layer. In particu-
lar, the first linear layer L
20
is a fully connected layer taking the 12-vector vto 20 neurons by
simply the linear function y=wx +b. This is then fed into an element-wise sigmoid layer
sxð Þ ¼ 1þex
ð Þ1of 20 neurons, followed again by a linear layer, which is then summed to
the real number �as the output. The schematic of is shown in Fig 2. We have taken this neu-
ral network only to illustrate the power of our methodology and have not optimised the
hyper-parameters such as 20, nor the network architecture or the choice of the type of
neurons.
Fig 2. The structure of the neural network used in the calculation.
https://doi.org/10.1371/journal.pone.0250227.g002
PLOS ONE
Machine-learning a virus assembly fitness landscape
PLOS ONE | https://doi.org/10.1371/journal.pone.0250227 May 5, 2021 5 / 10
• Now split Dinto a training set Tof 30,000 random samples; the validation set will be the
complement V=D\T. Note that the training data is only about 5.6% of the total data.
• We train with Tand validate on V.
• As a further check, we create a “fake” validation set ~
Vwhich has the same inputs as that of V
but with output randomly assigned from the set of correct outputs.
On an ordinary laptop (Intel Core m5-6Y57 CPU, 1.10GHz, 2 Cores, 4 Logical Processors,
with 8Gb of RAM), the training took about 45 seconds, and the prediction about 10 seconds.
The algorithm is implemented on Mathematica [23] and is expected to run even faster on the
Python package Tensorflow [24]. A python Jupyter notebook is attached along with the
Mathematica notebook as S1 and S2 Files. In other words, the entire computation took
under 1 minute with an ordinary notebook as opposed to the many hours it took on a super-
computer. As mentioned above, a number of similar neural networks were all able to per-
form this supervised learning task in a comparably short time, meaning there was little moti-
vation to optimise neural network architecture or hyperparameters. However, this reinforces
the point that there is intrinsic structure in the costly simulation dataset that any reasonable
neural networks finds easy to learn. We also tried various combinations of other standard
machine-learning algorithms such as decision trees and support vector machines, and
empirically find that our particular neural network approach above seems to out-perform all
of them.
We present the result in Fig 3. In part A, we present a plot of the predicted �on the horizon-
tal versus the actual �on the vertical. There are 501, 441 points. One can see that they cluster
near the desired y=xline, which would mean perfect prediction (note the axis ranges). To
give some precise measures, the best fit line is y=−0.0262122 + 1.02519xwith F-statistic
2.45082 ×10
6
and p-value less than 10106. The R-squared value is 0.830151. The errors them-
selves (fit-residuals) give a mean and standard deviation of (6.88 ±0.02) ×10
−16
, showing that
the residuals are unbiased around 0. To double check, we plot the same result for the fake vali-
dation set ~
Vin part B. It is obvious that the distribution is much less structured and essentially
randomly occupies a square. The fit here is y= 0.815233 −0.000947474xi.e. practically a con-
stant, with a poor F-statistic of 0.450401 and a poor p-value of 0.502145. The R-squared value
is 8.98217 ×10
-7
. This is very re-assuring for less than 6% of seen data and total computation
time of less than 1 minute on an ordinary laptop.
To give an idea of the prediction, for the (1, 1, . . ., 1) vector, the net predicts �= 0.87069, or
1741. The original value is 200, but that is a singular outlier in the whole data set, which we
would not expect the neural network to be able to reproduce. For the (2, 2, . . ., 2) vector, the
net predicts �= 0.834721, or 1669; the correct value is 1745. For the (3, 3, . . ., 3) vector, the net
predicts �= 0.673568, or 1347; the correct value is 1309.
We will use R-squared, a real number between 0 and 1, as a measure of accuracy of the
machine-learning; the closer it is to 1, the better the fit (for a good reference on machine-learn-
ing and goodness of fit measure, cf. e.g. [25]). Our 30, 000 training set was only to illustrate the
technique in detail. In general, we need to perform cross-validation by splitting the dataset.
We split the data into a fraction x of random samples for training and validate on the comple-
ment 1 −x, done for training set from 30, 000 to 500, 000, in steps of 30, 000. The R-squared
value is computed for each case as a measure of precision. Moreover, for each x, we repeat the
random sampling 10 times, for which we get the error bars. The plot of the R-squared (with
error bars) against the increase of percentage x of the size of the training set con-stitutes a
learning curve which illustrates how the neural network responds to the data; this is shown in
Fig 4.
PLOS ONE
Machine-learning a virus assembly fitness landscape
PLOS ONE | https://doi.org/10.1371/journal.pone.0250227 May 5, 2021 6 / 10
Fig 3. A A scatter plot of the predicted and actual value of �for the validation data from a 30, 000 training sample; Bscatter plot of a random-prediction
versus the actual �value.
https://doi.org/10.1371/journal.pone.0250227.g003
PLOS ONE
Machine-learning a virus assembly fitness landscape
PLOS ONE | https://doi.org/10.1371/journal.pone.0250227 May 5, 2021 7 / 10
As a comparison, one might imagine that since there is an underlying pattern being
machine-learnt, a simple regression might suffice. That is, could one fit a hyperplane fðxiÞ ¼
a0þP
12
i¼1
aivito the data? We perform this over the entire dataset, and find that the best multi-
linear regression obtains only R
2
= 0.575428. Introducing non-linearity and more parameters,
such as fitting fðxiÞ ¼ a0þP
12
i¼1
aiviþP
12
i¼1
biv2
idoes not do much better, at R
2
= 0.665484. The
inherent complexity (non-linearity) of the problem is therefore best captured by our neural
network approach.
Supporting information
S1 File. Mathematica notebook implementing the neural network described in the article.
(NB)
S2 File. Python Jupyter notebook implementing similar neural networks.
(IPYNB)
Acknowledgments
We would like to thank Richard Bingham and Eric Dykeman for making available their code,
fitness landscape and figures. We also thank Reidun Twarock, Richard Bingham and Eric
Dykeman for interesting discussions. We would both like to express our gratitude to the Insti-
tute for Computational and Experimental Research in Mathematics in Providence, Rhode
Island, where part of this work was carried out, for their hospitality.
Author Contributions
Conceptualization: Pierre-Philippe Dechant.
Data curation: Pierre-Philippe Dechant.
Formal analysis: Pierre-Philippe Dechant, Yang-Hui He.
Fig 4. Learning curve for the R-square value versus fraction (seen) training data from 0 to 1.
https://doi.org/10.1371/journal.pone.0250227.g004
PLOS ONE
Machine-learning a virus assembly fitness landscape
PLOS ONE | https://doi.org/10.1371/journal.pone.0250227 May 5, 2021 8 / 10
Investigation: Pierre-Philippe Dechant.
Methodology: Pierre-Philippe Dechant, Yang-Hui He.
Project administration: Pierre-Philippe Dechant.
Resources: Pierre-Philippe Dechant.
Software: Yang-Hui He.
Validation: Pierre-Philippe Dechant, Yang-Hui He.
Visualization: Pierre-Philippe Dechant, Yang-Hui He.
Writing – original draft: Pierre-Philippe Dechant, Yang-Hui He.
Writing – review & editing: Pierre-Philippe Dechant, Yang-Hui He.
References
1. Twarock R, Stockley PG. RNA-mediated virus assembly: mechanisms and consequences for viral evo-
lution and therapy. Annual Review of Biophysics. 2019;. https://doi.org/10.1146/annurev-biophys-
052118-115611 PMID: 30951648
2. Dechant PP, Twarock R, et al. Models of viral capsid symmetry as a driver of discovery in virology and
nanotechnology. The Biochemist. 2021; 43(1):20–24. https://doi.org/10.1042/bio_2020_102
3. Stockley PG, White SJ, Dykeman E, Manfield I, Rolfsson O, Patel N, et al. Bacteriophage MS2 genomic
RNA encodes an assembly instruction manual for its capsid. Bacteriophage. 2016; 6(1):e1157666.
https://doi.org/10.1080/21597081.2016.1157666 PMID: 27144089
4. Koning RI, Gomez-Blanco J, Akopjana I, Vargas J, Kazaks A, Tars K, et al. Asymmetric cryo-EM recon-
struction of phage MS2 reveals genome structure in situ. Nature communications. 2016; 7. https://doi.
org/10.1038/ncomms12524 PMID: 27561669
5. Patel N, Dykeman EC, Coutts RH, Lomonossoff GP, Rowlands DJ, Phillips SE, et al. Revealing the den-
sity of encoded functions in a viral RNA. Proceedings of the National Academy of Sciences. 2015; 112
(7):2227–2232. https://doi.org/10.1073/pnas.1420812112 PMID: 25646435
6. Twarock R, Leonov G, Stockley PG. Hamiltonian path analysis of viral genomes. Nature communica-
tions. 2018; 9(1):2021. https://doi.org/10.1038/s41467-018-03713-y PMID: 29789600
7. Patel N, Wroblewski E, Leonov G, Phillips SE, Tuma R, Twarock R, et al. Rewriting nature’s assembly
manual for a ssRNA virus. Proceedings of the National Academy of Sciences. 2017; p. 201706951.
https://doi.org/10.1073/pnas.1706951114 PMID: 29087310
8. Zlotnick A. To build a virus capsid: an equilibrium model of the self assembly of polyhedral protein com-
plexes. Journal of molecular biology. 1994; 241(1):59–67. https://doi.org/10.1006/jmbi.1994.1473
9. Bingham RJ, Dykeman EC, Twarock R. RNA Virus Evolution via a Quasispecies-Based Model Reveals
a Drug Target with a High Barrier to Resistance. Viruses. 2017; 9(11):347. https://doi.org/10.3390/
v9110347
10. Gillespie DT. Exact stochastic simulation of coupled chemical reactions. The journal of physical chemis-
try. 1977; 81(25):2340–2361. https://doi.org/10.1021/j100540a008
11. Dykeman EC. A Model for Viral Assembly around an Explicit RNA Sequence Generates an Implicit Fit-
ness Landscape. Biophysical Journal. 2017; 113(3):506–516. https://doi.org/10.1016/j.bpj.2017.06.037
PMID: 28793206
12. Singh AR, Kos
ˇmrlj A, Bruinsma R. Finite temperature phase behavior of viral capsids as oriented parti-
cle shells. Physical review letters. 2020; 124(15):158101. https://doi.org/10.1103/PhysRevLett.124.
158101 PMID: 32357054
13. Li S, Orland H, Zandi R. Self consistent field theory of virus assembly. Journal of Physics: Condensed
Matter. 2018; 30(14):144002. https://doi.org/10.1088/1361-648X/aab0c6 PMID: 29460850
14. Twarock R, Bingham RJ, Dykeman EC, Stockley PG. A modelling paradigm for RNA virus assembly.
Current Opinion in Virology. 2018; 31:74–81. https://doi.org/10.1016/j.coviro.2018.07.003 PMID:
30078702
15. Eigen M. Viruses: evolution, propagation, and defense. Nutrition reviews. 2000; 58(s1). PMID:
10748612
16. Lauring AS, Andino R. Quasispecies theory and the behavior of RNA viruses. PLoS pathogens. 2010; 6
(7). https://doi.org/10.1371/journal.ppat.1001005 PMID: 20661479
PLOS ONE
Machine-learning a virus assembly fitness landscape
PLOS ONE | https://doi.org/10.1371/journal.pone.0250227 May 5, 2021 9 / 10
17. Lauring AS, Andino R. Exploring the fitness landscape of an RNA virus by using a universal barcode
microarray. Journal of virology. 2011; 85(8):3780–3791. https://doi.org/10.1128/JVI.02217-10 PMID:
21289109
18. Degiacomi MT. Coupling Molecular Dynamics and Deep Learning to Mine Protein Conformational
Space. Available at SSRN 3213915. 2018.
19. Hastie T, Tibshirani R, Friedman J. The elements of statistical learning: data mining, inference, and pre-
diction. Springer Science & Business Media; 2009.
20. He YH. Deep-learning the landscape. arXiv preprint arXiv:170602714. 2017.
21. He YH. Machine-learning the string landscape. Phys Lett. 2017; B774:564–568. https://doi.org/10.
1016/j.physletb.2017.10.024
22. He YH. The Calabi-Yau Landscape: from Geometry, to Physics, to Machine-Learning. to appear in Lec-
ture Notes in Mathematics, Springer; 2021.
23. Wolfram Research Inc. Mathematica, Version 11.3.
24. Martı
´n Abadi et al. TensorFlow: Large-Scale Machine Learning on HeterogeneousSystems;. Available
from: http://tensorflow.org/.
25. Haykin S. Neural Networks: A Comprehensive Foundation. 1st ed. Upper Saddle River, NJ, USA:
Prentice Hall PTR; 1994.
PLOS ONE
Machine-learning a virus assembly fitness landscape
PLOS ONE | https://doi.org/10.1371/journal.pone.0250227 May 5, 2021 10 / 10
