When proxy-tissue QTLs work and when they mislead, per geneticist Sasha Gusev
SashaGusevPosts · x · 2026-08-29
In a thread with Anshul Kundaje, Sasha Gusev outlines methodology: for genes/elements active in relevant tissues, cross-tissue genetic correlation of QTLs is likely high, so proxy tissues (even seemingly irrelevant ones) give similar results to the "right" tissue for QTL mapping or TWAS/coloc. But genes completely inactive in relevant tissues are strongly depleted of trait heritability—don't use those proxy QTLs, e.g., avoid blood QTLs from brain-off genes when studying schizophrenia (simplified).
More from Research
- AI doesn't mean the end of mathematics – yet — ArtificialOther · 2026-08-29
- Multi-agent science world writes 125 cited papers, finds new 604-sphere record beating AlphaEvolve — progenitor414 · 2026-08-29
- Using Datalog Engine to Fix LLM Memory Inconsistency — petrusenko_max · 2026-08-29
- Debate on CoT: Anthropomorphism obscures technical limitations in LLM reasoning — rao2z · 2026-08-29
- EdgeBench paper reveals log-sigmoid scaling laws for agent learning — rohanpaul_ai · 2026-08-29
- Validation and Benchmarking are the most important skills — A_K_Nain · 2026-08-29