Deep Learning for Multi-Omic Integration
I develop interpretable methods for multi-omic data integration that connect molecular layers such as gene expression, epigenomics, and proteomics. The goal is to understand how these layers jointly shape health and disease.
Status Ongoing project
Selected publications:
- Unified imputation of missing data modalities and features in multi-omic data via shared representation learning (bioRxiv, 2026)
Collaborators: William Stafford Noble , MOHD Consortium
Protein Language Models and Protein Science
I develop and analyze protein language models (PLMs) to understand how evolutionary information is encoded in protein sequences. My work spans transformer architectures, specialized models for intrinsically disordered proteins, and studies showing that PLMs capture epistasis, fitness landscapes, and evolutionary constraints.
Status Ongoing project
Selected publications:
- Protein Language Models Capture Structural and Functional Epistasis in a Zero-Shot Setting (bioRxiv, 2025)
- DR-BERT: A Protein Language Model to Annotate Disordered Regions (Structure, 2024)
- Transforming the Language of Life: Transformer Neural Networks for Protein Prediction Tasks (ACM BCB, 2020)
- Machine Learning for Protein Science and Engineering (Cold Spring Harbor Perspectives in Biology, 2025)
Collaborators: Sergei Maslov, Mark Hopkins, Anna Ritz, Shivani Ahuja, Greg Anderson
Transcriptional Regulation and Functional Genomics
I build models linking promoter sequence, transcription-factor activity, and chromatin context to gene-expression programs. In FUN-PROSE, we predicted condition-specific expression in fungi from sequence and TF abundance, revealing interpretable motif–TF relationships. I also contributed to a transcriptomic atlas of acid-stress response in multiple Issatchenkia orientalis strains.
Status Ongoing project
Selected publications:
- FUN-PROSE: A Deep Learning Approach to Predict Condition-Specific Gene Expression in Fungi (PLoS Comput Biol, 2023)
- A Transcriptomic Atlas of Acute Stress Response to Low pH in Multiple Issatchenkia orientalis Strains (Microbiology Spectrum, 2024)
Collaborators: Sergei Maslov, William Stafford Noble
Machine-Learning-Guided GWAS and Disease Prediction
I apply machine learning to enhance genome-wide association studies, integrating clinical, environmental, and genetic data to uncover patterns underlying complex diseases.
Status Ongoing project
Selected publications:
- ML-Guided GWAS Reveals Genetic Architectures for MASLD for Overweight and Lean Individuals in the All of Us Cohort (medRxiv, 2025)
Collaborators: Sergei Maslov, Sharon Donovan, Konstantinos Lazaridis, Alina Allen
Information Theory for Diagnostics
I applied ideas from information theory to pathogen diagnosis. Our semi-quantitative group testing framework enables efficient, accurate two-stage qPCR screening across wide load ranges while preserving quantitative resolution.
Selected publication:
- Semi-Quantitative Group Testing for Efficient and Accurate qPCR Screening of Pathogens with a Wide Range of Loads (BMC Bioinformatics, 2024)
Collaborators: Sergei Maslov, Olgica Milenkovic
Evolution and Classification of Technology
I investigate innovation dynamics, diversity, and recombination in technological ecosystems using patent and product datasets. We modeled how novelty sustains open-ended innovation and quantified a long-term decline in product diversity among large U.S. firms.
Selected publications:
- Open-Ended Technological Innovation (Artificial Life, 2019)
- Dropping Diversity of Products of Large US Firms: Models and Measures (PLOS ONE, 2022)
Collaborators: Mark Bedau, Norman Packard, Tobias Rubel, J. Doyne Farmer
