Philip John FredaCedars-Sinai Medical Center
Philip John Freda
Doctor of Philosophy
Development of machine learning algorithms and pipelines for the analysis of complex traits
About
23
Publications
42,547
Reads
How we measure 'reads'
A 'read' is counted each time someone views a publication summary (such as the title, abstract, and list of authors), clicks on a figure, or views or downloads the full-text. Learn more
180
Citations
Introduction
I am a postdoctoral scientist investigating the basis of complex traits using machine learning and artificial intelligence at Cedars-Sinai Medical Center.
I am broadly interested in the genetic and environmental features of complex traits in many systems and phenotypes. My interests range from psychiatric disease in humans to the detection of non-additive genetic effects across systems and phenotypes.
Additional affiliations
January 2022 - present
Position
- Postdoctoral Scientist
Description
- Research in complex traits in humans, mice, and Drosophila focusing using artificial intelligence and machine learning.
May 2019 - January 2022
Position
- PostDoc Position
Description
- Research in the risk assessment of the development of opioid use disorder (OUD)
August 2016 - December 2016
Position
- Lecturer
Description
- Creation and instruction of lectures focusing on the relationships between insects and humans. Lectured on insect biology, evolution, taxonomy, behavior, and ecology. Grading of exams, essays, and quizzes
Education
August 2014 - December 2018
August 2012 - May 2014
August 2001 - December 2005
Publications
Publications (23)
As organisms age the environment fluctuates, exerting differential selection across ontogeny. In particular, highly seasonal environments expose life stages to often drastically different thermal environments. This developmental variation is particularly striking in organisms with complex life cycles, wherein life history stages also exhibit distin...
Overall, our results provide substantial evidence that thermal hardiness and developmental acclimation responses are decoupled across metamorphosis in D. melanogaster. We do provide evidence that environmental variation can substantially alter genetic correlations among traits (heat hardiness and cold hardiness), but not genetic correlations across...
Organisms with complex life cycles demonstrate a remarkable ability to change their phenotypes across development, presumably as an evolutionary adaptation to developmentally variable environments. Developmental variation in environmentally sensitive performance, and thermal sensitivity in particular, has been well documented in holometabolous inse...
The opioid epidemic continues to contribute to loss of life through overdose and significant social and economic burdens. Many individuals who develop problematic opioid use (POU) do so after being exposed to prescribed opioid analgesics. Therefore, it is important to accurately identify and classify risk factors for POU. In this review, we discuss...
Background
Quantitative Trait Locus (QTL) analysis and Genome-Wide Association Studies (GWAS) have the power to identify variants that capture significant levels of phenotypic variance in complex traits. However, effort and time are required to select the best methods and optimize parameters and pre-processing steps. Although machine learning appro...
Statistical epistasis has been studied extensively because of its potential to provide evidence for genetic interactions for phenotypes, but there have been methodological limitations to its exhaustive, widespread application. We present new algorithms for the interaction coefficients for standard regression models for epistasis that permit many va...
Our objective was to detect common barriers to post-acute care (B2PAC) among hospitalized older adults using natural language processing (NLP) of clinical notes from patients discharged home when a clinical decision support system recommended post-acute care. We annotated B2PAC sentences from discharge planning notes and developed an NLP classifier...
Background
Quantitative Trait Locus (QTL) analysis and Genome-Wide Association Studies (GWAS) have the power to identify variants that capture significant levels of phenotypic variance in complex traits. However, effort and time are required to select the best methods and optimize parameters and pre-processing steps. Although machine learning appro...
Background
Opioid use disorder (OUD) is underdiagnosed in health system settings, limiting research on OUD using electronic health records (EHRs). Medical encounter notes can enrich structured EHR data with documented signs and symptoms of OUD and social risks and behaviors. To capture this information at scale, natural language processing (NLP) to...
Organisms with complex life cycles demonstrate a remarkable ability to change their phenotypes across development, presumably as an evolutionary adaptation to developmentally variable environments. Developmental variation in environmentally sensitive performance, and thermal sensitivity in particular, has been well documented in holometabolous inse...
Opioid use disorder (OUD) creates significant public health and economic burdens worldwide. Therefore, understanding the risk factors that lead to the development of OUD is fundamental to reducing both its prevalence and its impact. Significant sources of OUD risk include co-occurring lifetime and current diagnoses of both psychiatric disorders, pr...
Environmental challenges presented by temperature variation can be overcome through phenotypic plasticity in small invasive ectotherms. We tested the effect of thermal exposure to 21, 18, and 11°C throughout the whole life cycle of individuals, thermal exposure of adults reared at 25°C to 15 and 11°C for a 21-d period, and long (14:10 hr) and short...
The adaptive decoupling hypothesis - complex life cycles evolve to genetically separate life stages, allowing stage-specific responses to selective forces. Insects that undergo complete metamorphosis are often subject to distinct thermal habitats, resulting in differential selection across development. We test for genetic decoupling in thermal hard...
A presentation explaining how and why larvae and adults may reach distinct fitness optima due to a genetic decoupling of thermal stress responses. Takes about possibly differences in RNAseq profiles
Does decoupled physiology via a complex life cycle lead to a decoupled stress response and potential evolutionary independence across metamorphosis? This presentation points to yes!
Showing data of phenotypic correlation for cold hardiness between larvae and adults of Drosophila melanogaster. There are 139 line used in this study. There are also manhattan plots showing SNPs that are in association with variation in cold hardiness in each life stage. From a 3 minute thesis competition
Presentation at SICB 2016 discussing my findings of a complete phenotype and quantitative genetic decoupling of thermal hardiness across metamorphosis in Drosophila melanogaster
A proposal talk discussing the possibility and implications of complete genetic decoupling of stress traits across metamorphosis with preliminary results.
Poster showing the decoupling of cold hardiness across metamorphosis at both a phenotypic and quantitative level in Drosophila melanogaster
A presentation talking about two main points:
1.) If complex life cycles evolved to decouple life stages, does the underlying genetic architecture associated with a trait change across ontogeny? If so, to what degree?
and
2.) What are the physiological mechanisms facilitating observed changes?
The Muller F element (4.2 Mb, ~80 protein-coding genes) is an unusual autosome of Drosophila melanogaster; it is mostly heterochromatic with a low recombination rate. To investigate how these properties impact the evolution of repeats and genes, we manually improved the sequence and annotated the genes on the D. erecta, D. mojavensis, and D. grimsh...
A good Drosophila trap should be made of materials that are inexpensive and readily available. Also, the materials should be sturdy enough to be used outdoors. Additionally, a trap should be simple enough that anyone can assemble it quickly. The trap of Medeiros and Klaczko (1999) is well designed, but improvements and simplifications are possible....
Collections of Drosophila and their relatives were performed using bait traps on the campus of Saint Joseph's University, in Philadelphia and Lower Merion, Pennsylvania, between July and December 2011 and continuing in March of 2012. In the 2011 collection season, more than 200 specimens of Drosophila suzukii (Matsumura), or spotted wing Drosophila...
Questions
Questions (2)
I have sets of FlyBase gene ID's that I would like to perform KEGG Pathway analysis on but the KAAS (http://www.genome.jp/kaas-bin/kaas_main) seems to require sequence data to perform BLAST with. Is there a way that I can just submit my gene ID's and perform the analysis that way?
I have been working with a QTL dataset for quite some time and keep stumbling over an issue. Out of the 1000 or so markers I have, a number of them have the same location value (in cM) as other markers.
In other words:
marker x: 10.05 cM
marker y: 10.05 cM
marker z: 10.05 cM
First, I removed all of the markers that had some type of overlap associated, leaving only one that was associated with some physical gene or bp location. This limited my dataset to approximately 215 markers, a far cry from my original 1000.
I got the results and then tried it again with the original 1000 markers. I am using QTL Cartographer (Unix) v. 1.17 for the analysis. Interestingly, during the analysis in both LRmapqtl and SRmapqtl, I noticed strange output within the terminal like "cond.denom=0" over and over again. I am guessing this is occurring because the recombinational distances between some markers is 0 and therefore are being excluded. However, the output from the 1000 marker analysis yielded more QTL with higher resolution with peaks being in the same general locations as the 215 marker analysis (some just became significant in the 1000 marker analysis)
Should I remove the markers and stick with the ~215 that do not overlap?
Will my data be erroneous if I leave the other ~800 or so markers in?
Any help would be appreciated.
