Gowri Nayar
Ph.D. Candidate in Biomedical Informatics, Stanford University
Hi! I’m a Ph.D. candidate at Stanford University, where I’m advised by Prof. Russ Altman.
My research focuses on understanding protein function and biological pathways through machine learning. I develop graph neural networks and protein language model methods, along with statistical benchmarks, with a focus on the understudied proteome. My goal is to build scalable AI systems that turn large-scale proteomic and genomic data into biological insight, particularly for predicting novel protein pathways related to disease.
Previously, I was a machine learning team lead at IBM Research, where I designed large-scale genomic analysis systems and worked on NLP and deep learning for biological sequence modeling.
You can download my CV here.
Honors: NIH F31 Fellowship (2024–2026) · Stanford Data Science Fellowship · Donald V. Jackson Fellowship
news
| Jul 08, 2026 | Our paper GATSBI: Improving context-aware protein embeddings through biologically motivated data splits was presented at ISMB 2026 and appears in Bioinformatics.  |
|---|---|
| Sep 15, 2025 | New paper out in PLOS Computational Biology: Paying attention to attention: High attention sites as indicators of protein family and function in language models. |
| Jun 01, 2025 | Pool PaRTI, our PageRank-based pooling method for protein sequence representations, is published in Bioinformatics. Code is available on GitHub. |
selected publications
- Heterogeneous network approaches to protein pathway predictionComputational and Structural Biotechnology Journal, 2024
- Computational Approaches to Drug Repurposing: Methods, Challenges, and OpportunitiesAnnual Review of Biomedical Data Science, 2024
- Paying attention to attention: High attention sites as indicators of protein family and function in language modelsPLOS Computational Biology, 2025
- GATSBI: Improving context-aware protein embeddings through biologically motivated data splitsBioinformatics, 2026ISMB 2026