Blog

Five principles from Dixit and Nalebuff's The Art of Strategy and how they already show up in genomic data infrastructure, research bets and model trust.

Four scientific workbenches, one open skill standard and the validation layer nobody has built yet, the long read behind Sorting My DNA, Issue 1.

A simple classification task, two approaches, and the coefficient problem hiding behind the spend-more-tokens culture.

Four books. Centuries apart. The same framework for navigating a world that won't stop moving.

Methods companion to the TERT promoter post: notebook setup, gnomAD control selection, model choice, AlphaGenome API calls, score_variant caveats and evaluation against published biology.

A deep-dive into TERT promoter mutations using AlphaGenome to reconstruct chromatin, TF binding and transcriptional effects from sequence alone.

A practical tool to check whether your tissue or cell type is covered in AlphaGenome training tracks before you run variant effect analyses.

U-Net DNA, teacher–student distillation, 5,930 tracks, GWAS to T-ALL: how AlphaGenome compresses a megabase, attends globally and scores variants, plus how to run it.

From a missing pancreas to cis-regulatory logic: why most disease variants sit outside genes and the range-versus-resolution problem AlphaGenome was built to solve.

A data-driven analysis of AI augmentation potential across computational biology and life sciences, using Anthropic's labor market data, 176 job descriptions and a live test on my own role.

What I learned building agents for the past two years. A practical guide to the frameworks, tools and patterns that actually work.

I'm excited to share insights about AI, genomics and research. This is the beginning of a journey to document my learnings and discoveries.