Research

Research in the Moses Lab

The research projects we pursue typically weave together many threads from disciplines such as evolutionary genetics, systems biology, machine learning, sequence analysis, computer vision, and more. The projects we have pursued over the years reflect the diverse interests of the graduate students and postdocs who worked in the lab. Below is a listing of the main themes that define and guide many of our research projects. For a complete picture of the work that we’ve accomplished, please visit our Publications page.

Evolution and Dynamics of Regulatory Networks

CRZ1 Pulse - Ian Hsu
Zhang & Mangelsdorf 2002
Kompella et al. 2017
previous arrowprevious arrow
next arrownext arrow
Slider

Most complex cellular processes are carried out by groups of genes working together in so-called pathways or networks. We seek to understand how these networks are encoded in genome sequences, how they create dynamic biological phenotypes, and they are created by evolution.

LATEST PAPERS:

Mehdi TF, Singh G, Mitchell JA, Moses AM. Variational Infinite Heterogeneous Mixture Model for Semi-supervised Clustering of Heart Enhancers. Bioinformatics 2019 Feb 7
Bioinformatics link PDF Mirror PubMed Abstract Article
Regulatory networks Transcription factors Genomics Machine Learning
Hsu IS, Strome B, Plotnikov S, Moses AM. A Noisy Analog-to-Digital Converter Connects Cytosolic Calcium Bursts to Transcription Factor Nuclear Localization Pulses in Yeast. G3 2019 Feb 7;9(2):561-570.
G3 link PDF Mirror PubMed Abstract Article
Regulatory networks Subcellular localization Time lapse microscopy Cell signaling

Microscope images are big data

Lu et al. 2016
Handfield et al. 2015
previous arrowprevious arrow
next arrownext arrow
Slider

Automated microscopy has made it possible to measuring protein abundance and subcellular localization in millions of single cells. We are developing computational tools to extract basic biology from huge collections of microscope images without have to look at each one.

LATEST PAPERS:

Lu AX, Lu AX, Schormann W, Ghassemi M, Andrews DW, Moses AM The Cells Out of Sample (COOS) dataset and benchmarks for measuring out-of-sample generalization of image classifiers Advances in Neural Information Processing Systems 32 (NeurIPS 2019)
NeurIPS link PDF Mirror Book Chapter
Deep Learning Covariate shift Image analysis Subcellular localization
Lu AX, Kraus OZ, Cooper S, Moses AM. Learning unsupervised feature representations for single cell microscopy images with paired cell inpainting. PLoS Comput Biol. 2019 Sep 3;15(9):e1007348.
PLoS Comp. Bio link PDF Mirror PubMed Abstract Article
Deep Learning Self-Supervised Image analysis Subcellular localization

Molecular Evolution of Disordered Regions

Zarin et al. 2017
Zarin et al. 2017
Zarin et al. 2017
previous arrowprevious arrow
next arrownext arrow
Slider

Intrinsically Disordered Regions (or IDRs) are enigmatic protein regions that are involved in a wide variety of biological processes. Although they are widespread, they usually show little evolutionary conservation. Is this rapid evolution a sign that they are just "junk" protein, or do they facilitate evolutionary diversity? This question is also of medical relevance: when we find mutations in patients' IDRs we currently cannot tell what impact (if any) they are having.

LATEST PAPERS:

Pritišanac I, Vernon RM, Moses AM, Forman-Kay JD Entropy and Information within Intrinsically Disordered Protein Regions . Entropy 2019 21(7), 662
Entropy link PDF Mirror Review
Information theory Biophysics Intrinsically disordered Evolution
Zarin T Sequence-function relationships in intrinsically disordered regions through the lens of evolution.
University of Toronto 2019 Thesis Link PhD thesis
Molecular evolution Intrinsically disordered Sequence divergenceBiological function

Beautiful bioinformatics for genomics and proteomics

Davey et al. 2015
Lai et al. 2012
previous arrowprevious arrow
next arrownext arrow
Slider

Complete sequencing of genomes is now routine, and yields thousands of genes and proteins, and information about the genetic differences in populations. All of this data needs to be organized and analyzed: bioinformatics!

LATEST PAPERS:

Mehdi TF, Singh G, Mitchell JA, Moses AM. Variational Infinite Heterogeneous Mixture Model for Semi-supervised Clustering of Heart Enhancers. Bioinformatics 2019 Feb 7
Bioinformatics link PDF Mirror PubMed Abstract Article
Regulatory networks Transcription factors Genomics Machine Learning
Strome B, Hsu IS, Li Cheong Man M, Zarin T, Nguyen Ba A, Moses AM. Short linear motifs in intrinsically disordered regions modulate HOG signaling capacity. BMC Syst Biol. 2018 Jul 3;12(1):75.
BMC Systems Biology Link PDF Mirror PubMed AbstractArticle
Cell signaling Saccharomyces cerevisiae Mitogen-activated kinases Intrinsically disordered regions Short linear motifs High osmotic glycerol pathway