Alisa Omelchenko

About me

I'm a fifth-year PhD candidate in the CMU-Pitt Joint Computational Biology program, advised by Dr. Jishnu Das. My work sits at the intersection of machine learning and biology applying interaction-based language models, multi-omic network analysis, and protein modeling, in immunology and disease.

Before my PhD, I worked in drug discovery at a biotech startup and then as a lab manager in an academic neuroscience lab. It was in that academic setting that I recognized how computational tools and workflows can break down when they lose touch with the biology and the scientific audience they're meant to serve. That's what drove me to pursue a PhD. I wanted to build methods that are computationally rigorous and biologically relevant. Every benchmark, model, or pipeline I build gets tested against that standard.

Outside research, I mentor future scientists, advocate for science and education policy, and communicate scientific topics to the public.

What I'm focused on right now

Designing benchmarks and evaluation criteria that measure biological relevance
Building reproducible tools and pipelines that other researchers can use
Guiding agentic AI to explore testable hypotheses, and designing benchmarks for AI outputs to evaluate biological accuracy
Interaction language models (iLMs) for peptide, protein, and small-molecule interactions
Applying network and systems approaches to integrate biological information and fuel translational discovery in an unbiased manner
Patient-specific neo-epitope and variant prediction and prioritization in pediatric cancer

Research experience

St. Jude Center for Applied Bioinformatics · Summer Intern 2026 – Present

Predicted patient-specific cancer neo-epitopes and variants from a WGS dataset of a B-ALL cohort and identified patterns associated with relapse and subtypes for future clinical translation.
Directed agentic AI (Biomni and Co-Scientist) and other LLMs through prompt engineering to explore testable hypotheses and design validation strategies.
Evaluated AI outputs from Biomni and Co-Scientist for biological accuracy.
Mentored two high school interns in the Research Immersion Program, guiding them through hands-on research.

Das Lab, CMU-Pitt · Graduate Student 2022 – Present

Developed SWING, a first-in-class interaction-based language model to predict peptide:peptide interactions and perturbations (Nature Methods)
Integrated multi-omic datasets to identify a core Tfh cell network consistent across contexts (Nature Communications).
Collaborated to identify novel mechanisms of HIV infection in Tfh cells (In revision at Nature Communications).
Extending iLMs to capture ternary (pMHC:TCR) and non-peptide (small-molecule:protein) interactions for patient-specific, clinically translatable predictions (in progress).

Ni Lab, Virginia Tech · Laboratory Manager 2019 – 2022

Built an ImageJ plugin (TACI) to automate 3D calcium imaging analysis (JoVE).
Characterized the thermosensory pathway in Drosophila larvae (JEB, Front. Mol. Neurosci.).
Designed automated experimental workflows for fly and larval tracking to ensure reproducible results and speed up data analysis. (JEB)

Landos Biopharma · Laboratory Technician, Drug Discovery 2018 – 2019

Worked on target identification and mechanism-of-action studies by evaluating differentially expressed gene patterns.
Collected and analyzed data from necropsies, cell cultures, flow cytometry, ELISA, and western blots.

Download full CV (PDF) →

Skills

Machine learning

Interaction & protein language models (iLMs / pLMs), embedding-based downstream modeling, deeplearning (CNNs, transformers), interpretable ML.

Software & infrastructure

Python, R, Bash · HPC (Slurm, LSF) · published, adopted open-source tools.

Bioinformatics

Sequence analysis, protein structure modeling and visualization (AlphaFold3, PyMOL, OpenMM, fpocket), single-cell analysis (scRNA-seq, Multiome), multi-omic network analysis.

Biological databases

UniProt, PDB, Ensembl, NCBI, gnomAD, HINT, etc.

Wet lab

Genotyping/PCR, ELISA, flow cytometry, calcium imaging, confocal microscopy, molecular cloning, animal handling (mice, Drosophila).

Domain

Systems immunology, drug discovery (target identification, clinical data analysis), infectious diseases, pediatric oncology genomics, structural biology.

Selected publications

Nat
Comm2026

A unified network systems approach uncovers a core novel program underlying T follicular helper cell differentiation

Omelchenko AA, Rahman SA, Viswanadham VV, Yuen GJ, Estrada PMDR, D'Onofrio V, ... Das J.

Nat Comm Nature Communications

Nat
Meth2025

Sliding Window Interaction Grammar (SWING): a generalized interaction language model for peptide and protein interactions

Omelchenko AA*, Siwek JC*, Chhibbar P*, Arshad S, Rosengart A, Nazarali I, ... Das J.

Nature Methods, 1–13.

Res.
Square2026

IL-10- and TGF-β-driven BCL6 expression suppresses antiviral defenses and renders lymph node T follicular helper cells permissive to HIV infection

Ribeiro SP*, Del Rio Estrada PM*, Georgakis S, Orfanakis M, Omelchenko A, Coirada F, Dos Santos J, et al.

Preprint In revision, Nature Communications

JoVE2022

TACI: An ImageJ Plugin for 3D Calcium Imaging Analysis

Omelchenko AA, Bai H, Hussain S, Tyrrell JJ, Klein M, Ni L.

Journal of Visualized Experiments, (190), e64953.

FMN2022

Cool and warm ionotropic receptors control multiple thermotaxes in Drosophila larvae

Omelchenko AA, Bai H, Spina EC, Tyrrell JJ, Wilbourne JT, Ni L.

Frontiers in Molecular Neuroscience.

JEB2022

Responses of different Drosophila species to temperature changes

Huda A*, Omelchenko AA*, Vaden TJ*, Castaneda AN, Ni L.

Journal of Experimental Biology, 225(11), jeb243708.

* denotes equal contribution. Full list on Google Scholar.

Get in touch

I'm always happy to talk about interaction language models, immunology, science policy/communication, or the state of tools in computational biology. Feel free to reach out about potential collaborations, questions on my research, or just to say hi.

alisa.a.omelchenko@gmail.com

Hi, I'm Alisa 👋