Review-ready finding
KDM1A — structured candidate finding
Claim
KDM1A passed all registered source-QC gates in Rest, 8 h stimulation, and 48 h stimulation and ranked #6 by its weakest condition priority score. It is a reproducible source-QC candidate for independent follow-up, not an experimentally validated target.
Supporting evidence
- 48 h stimulation: rank #4, priority score 0.913, minimum donor concordance 0.718, and guide concordance 0.813.
- 8 h stimulation: rank #5, priority score 0.900, minimum donor concordance 0.692, and guide concordance 0.844.
- Rest: rank #14, priority score 0.851, minimum donor concordance 0.629, and guide concordance 0.799.
- Cross-condition maximin summary: rank #6, minimum score 0.851, with Rest as the lowest-scoring condition.
- The author-provided descriptive IL-10/IL-21 follow-up contains 5 observed donors for KDM1A; this is contextual follow-up, not independent confirmation.
Limitations
- The priority score is a relative triage heuristic, not a probability, p-value, or effect size.
- Downstream discovery count is power-sensitive, and donor/guide concordance values were reported by the source authors.
- Condition rank differences are descriptive because this workflow did not fit a perturbation-by-condition interaction model.
Evidence gap
Independent replication, orthogonal perturbation, formal endpoint testing, mechanism, safety, and therapeutic utility remain unestablished.
Recommended next experiment
Run an independent multi-donor CRISPRi replication for KDM1A with a prespecified functional endpoint in Rest, 8 h, and 48 h conditions; fit a formal perturbation-by-condition interaction before making a condition-specific claim, then confirm with an orthogonal perturbation.