Skip to content

overlake.bio

I built Spacefarer Phenome, a genetics-first computational drug-discovery pipeline for spaceflight physiological stress, anchored on bone mineral density (eBMD).

Core trick

Runs a nine-stage funnel with Mendelian randomization and PheWAS safety gates on CPU before any GPU docking or molecular dynamics stage.

Tags

Description

I built Spacefarer Phenome, a genetics-first computational drug-discovery pipeline for spaceflight physiological stress, anchored on bone mineral density (eBMD). A nine-stage funnel (phenotype, GWAS, druggable-genome, Mendelian randomization, PheWAS safety, druggability, structure, docking, MD) spends cheap CPU on the causal and safety stages before any GPU. The headline is an honest negative: the drug-target MR scan validated no target in the pre-registered set of 33 — an instrument-limited result, since key pQTL/eQTL panels were unreachable at run time, not a biological null; the positive control passed strongly (IVW p=1.85e-44). Cathepsin K, a pre-specified exemplar (not discovered here), was carried end-to-end to exercise the structural arm, re-ranking 24 known ChEMBL inhibitors with a conditional MD result. What matters is the rigor: pre-registered thresholds frozen in git, JSON-schema contracts between stages, per-artifact provenance, and an openly licensed, auditable record. Total GPU cost: $41.06.
raimbetov/spacefarer-phenome-bwc★ 00 forksPythonApache-2.0Created Jul 8, 2026 · last push Jul 14, 2026

README excerpt

# SPACEFARER PHENOME

[Rakhan Aimbetov](https://orcid.org/0000-0003-0131-4530) for the **Built with Claude: Life Sciences** hackathon.

## Summary

A genetics-first computational drug-discovery pipeline that mines human genetic variation for targets relevant to spaceflight-induced physiological stress, then carries a validated small-molecule target through structure → docking → molecular dynamics stages.

The pipeline is tailored for execution using the Claude Science agentic harness, with full per-stage provenance.

**[`13_REPORTS/spacefarer_phenome_report.md`](13_REPORTS/spacefarer_phenome_report.md)** — the human-readable report (methods, results, limitations, safety caveats). Everything else supports it.

## Caveats

- The repository is an auditable record of an end-to-end pilot — a machine-readable blueprint, schema-validated stage-to-stage contracts, per-run provenance manifests, and the results output.
- **It is *not*:** a `clone && run` pipeline. Reproducing the numbers requires external data, credentials, and a specific execution runtime (see *Limitations* below).
- The pilot was executed with substantial AI assistance (bespoke specialist subagents orchestrating the funnel under the blueprint's rulings and gates). Every stage nonetheless emits human-inspectable artifacts, schema-validated contracts, and a provenance manifest.
- Treat AI-produced numbers as auditable, but to be independently confirmed.

## Key features

- **Enforced pre-registration** — statistical th

Team