Antidepressant Pharmacogenomics: Much More Than You Wanted To Know
Read the original on Slate Star Codex →
Summary
A model MMTYWTK evidence-review of genetically-guided antidepressant prescription (GeneSight). Lays out the appealing dream (pharmacogenomics could reconcile the clinician-vs-researcher antidepressant-efficacy dispute), then systematically dismantles it: the candidate genes (HTR2A, SLC6A4) rest on small/unreplicated studies; a GCTA shows massively-polygenic structure so a 7-gene panel can't do much; the warfarin cautionary tale (perfect candidate, still failed); dose-titration makes metabolism genes redundant; a detailed teardown of all 8 GeneSight studies (all company-run, unblinded/unrandomized, dropout-confounded); Carlat + Ioannidis authorities; and the sharp Section-VIII meta-point on skeptic-community epistemics. Tentatively rejects the edifice pending independent replication.
Why this score
Quality 77 · Excellent. Excellent: comprehensive, numerate, a model of medical-evidence evaluation with a genuine epistemic payoff (the 'what's the point of experts who just ignore inconvenient evidence' Section VIII), on the Marijuana-MMTYWTK tier. 77.
Claude’s paradigm shift 48 · Moderate. Rigorous evaluation, not a new paradigm — the value is in the skeptical synthesis. B48.
Real-world impact 2 · Minor. A practical consumer/clinician warning against a real $2000 test, influential within psychiatry discourse; no broad material reach. RWI2.