Ovchinnikov Lab · MIT Biology
from sequences to contacts to protein folding to structures to networks to genomes to evolution
News
-
June 2026 MSA Pairformer is now published in Cell!Our compact, MSA-based protein language model that captures the coevolution of interacting proteins — with open-source code and a Colab for quick experimentation.
We build a unified statistical model of protein evolution — integrating coevolution, phylogenetics, structure, and function — and use it to read the notes nature has been taking for billions of years.
Research →
A unified model of protein evolution, conformational landscapes, and metagenomic dark matter.
Publications →
Work on protein language models, structure prediction, design, and structure determination.
Tools →
Open-source software, including ColabFold and ColabDesign.
Join →
Openings for undergraduates, graduates, visiting students, and postdocs.