I’m a biostatistician and data scientist with an M.D. and a background in experimental neuropharmacology. I develop statistical analyses, data-validation pipelines, and automated reporting for clinical trials, and build machine-learning models for biological data, from protein-function prediction and single-cell perturbations to animal behavior. Beyond biomedical applications, I also work on algorithmic problems in scientific computing and autonomous navigation.
Experimental pharmacology and therapy of pain and behavioral pharmacology.
Featured publications
Biostatistics for clinical development.
Featured outputs
Collaborative scientific ML project on computation and pathfinding on Cayley and Schreier coset graphs.
Featured research outputs
Studying SLAM algorithms for home robot navigation.
Pavlov First Saint Petersburg State Medical University
Large-scale multi-label prediction of protein functions.
13th of 1,625 teams (top 0.8%)
Multi-output regression of gene-expression responses across cell types.
13th of 1,097 teams (top 1.2%)
Judges' Prize
Large-scale binary classification of compound-protein binding.
13th of 1,950 teams (top 0.7%)
Best student team solution
Mouse behavior classification from pose keypoint time series.
12th of 1,412 teams (top 0.8%)
Large-scale multi-label prediction of protein functions.
11th of 2,259 teams (top 0.5%, Solo)
GPU-accelerated PyTorch implementation of ML-guided beam search for permutation sorting; a spin-off from the CayleyPy research project.
Django app for managing and auditing abbreviations in .docx files; uses a fine-tuned Llama 3.2 model for medical-context suggestions.
Shiny app for clinical trial sample size calculation (currently bioequivalence; extensible to other designs).