Researcher working at the interface of neuroscience, natural products, and AI-driven drug discovery. My research focuses on computational modeling of neuroinflammation and neural plasticity, with an emphasis on identifying plant-derived and natural product-inspired compounds that may influence therapeutic pathways. I build in silico pipelines for molecular screening, receptor-state classification, and translational discovery workflows that connect computational biology with future wet-lab validation.
I am the author of Lacuna, an open-source tool for finding cryptic binding pockets that only appear as a protein moves. It is available on Claude Science, GPT-Rosalind, Tamarind Bio, and Neurosnap.
Featured
Cryptic binding pocket discovery via conformational ensemble analysis. Generates an ensemble, detects pockets in every conformer, and clusters them to surface sites that are closed in the static structure. MIT licensed, runs on a CPU, and ships an MCP server and Claude skill.
Lacuna is now usable across Anthropic’s Claude Science, OpenAI’s GPT-Rosalind, Tamarind Bio, and Neurosnap Inc. Lacuna is the open-source cryptic pocket tool I built to find binding sites that are invisible in a protein’s static structure and only appear as the protein moves. Instead of searching one conformation, it generates or accepts a conformational ensemble, detects pockets across every state,...
Lacuna, the open-source cryptic pocket tool I built, is now importable as a skill in Anthropic’s Claude Science. It finds cryptic binding pockets: sites that look closed or shallow in an unbound structure and only open when something binds. It generates a conformational ensemble, detects pockets in every conformer, and clusters them so a site that appears in only a...
Many disease-relevant targets are called undruggable not because they are biologically intractable, but because no binding site is visible in their ground state. K-Ras was considered undruggable for thirty years, until a cryptic pocket was found beneath its switch-II region. That pocket now backs sotorasib and adagrasib. Structure predictors return one static conformation, and most pocket finders search exactly that...
