Descriptions:
Nick Saraev walks through how he used Claude Code to build a personal genomic analysis pipeline that cross-references raw 23andMe DNA export data against publicly available genetic databases, producing a structured health report covering disease risk, supplementation recommendations, dietary guidance, and exercise protocols. The DNA test itself cost around $70–80 CAD; the AI-powered analysis added negligible cost on top, putting a process that previously required expensive specialist consultations within reach of anyone.
The system takes the raw genetic text file — a representation of roughly 3 billion base pairs encoded as sequences of A, T, C, and G — and uses Claude Code to query databases cataloging known SNP (single nucleotide polymorphism) associations. Saraev walks through his CLAUDE.md configuration file, which functions as a persistent high-level instruction injected at the start of every Claude Code session, and shares specific findings from his own results: he discovered he is a carrier for cystic fibrosis, identified a methylation pathway deficiency (MTHFR-related) corrected by methylfolate supplementation — reporting noticeable energy and sleep improvements within 2–3 days — and learned he metabolizes caffeine slowly, leading him to significantly reduce his intake.
The video also covers how to extend the pipeline further: generating custom meal plans tailored to genetic nutritional needs, adapting the system for family members, and building it out into a commercial health product. Saraev provides all templates and configuration files in the description, making the tutorial fully reproducible. It’s a compelling case study in using large language models for structured data analysis in a domain — personalized genomics — where accessibility has historically been gated by cost and specialist access.
📺 Source: Nick Saraev · Published January 26, 2026
🏷️ Format: Hands On Build
