MACHINE LEARNING & COMPUTATIONAL BIOLOGY
Machine learning for
questions in biology.
I build machine learning models that infer how genes regulate one another, and the evaluation that shows whether those inferences hold up.
Each node represents a gene. Connections represent possible regulatory relationships.
01 Publications
A. Hegde, K. Zook, Y. Wang, J. Cheng. “Multimodal graph reasoning for gene regulatory network inference.” NeurIPS 2026.
Co-author. I worked on the automated literature-validation agent and the multi-omic validation component, which combines chromatin-accessibility and transcriptomic data.
T. Nguyen, A. Hegde, K. Zook, J. Cheng. “GRNPred: A Multimodal Graph Transformer with Masked Gene Expression Pretraining for Gene Regulatory Network Inference” npj Systems Biology and Applications
Co-first author. I ran the full-model, ablation, and supplementary experiments for revision, fixed bugs in the training and evaluation code, and rewrote parts of the manuscript.
A. Hegde, K. Zook, J. Cheng. “Dynamic gene regulatory network inference.” (working title)
Co-author. Inferring gene regulatory networks that change over time, rather than treating them as fixed.
K. Zook, A. Hegde, J. Cheng. “A Benchmark for Evaluating Graph-Based Gene Regulatory Network Inference.”
First author. Planned for submission in December 2026.
02 Selected Work
01 / ML INFRASTRUCTURE
A self-hosted platform that runs ML experiments as queued jobs, built to survive repeated submissions and worker failures without corrupting results. Part of ongoing, unpublished research.
02 / GPU COMPUTING
Custom GPU kernels for the sparse message-passing operations used in graph neural networks. The goal was to reduce memory traffic and kernel-launch overhead during training.
03 / APPLIED AI
A pipeline that turns YouTube captions, metadata, and transcripts into searchable knowledge bases for individual creators, covering up to 4,000 videos per creator.
04 / SYSTEMS ENGINEERING
An in-memory key-value store built from the socket layer upward, including custom memory management and a TCP request path for real-time SET and GET operations.
03 About
I’m a junior at the University of Missouri–Columbia studying computer science through the accelerated B.S./M.S. program, with minors in mathematics and biological sciences. I’m interested in the point where computation can make biological questions easier to test.
In Dr. Jianlin Cheng’s lab, I develop machine learning methods for gene regulatory network inference: learning, from genomic data, which genes control which, and how that changes over time. I developed the literature-verification module for EviGRN, which checks predicted regulatory links against published evidence (co-author, accepted to NeurIPS 2026). I am also co-first author on GRNPred, a GRN inference model now in review, where I ran the full and ablation experiments and helped revise the manuscript.
I also like systems work: job queues, GPU kernels, and retrieval pipelines. The less visible parts of a project matter to me—clean data, reproducible runs, and results that hold up when someone else tries them.
04 Research & Experience
05 Education
GPA 3.84 · High Dean's List, all semesters · Mark Twain Scholarship
Aug 2024 – May 2028
In progress
Mathematics & Biological Sciences