Proceedings of the National Academy of Sciences · 6 November 2026 · From issue 64
Bayesian knockoff filter for controlling false discovery rate
A new method finds real patterns in data with fewer false alarms than older tools.
Jiaqi Gu, Guosheng Yin et al.
A new statistical method helps scientists find the truly important factors hidden in huge piles of data. It avoids being fooled by random noise. The method builds many fake "decoy" versions of each factor inside a computer simulation, instead of making just one decoy. This keeps the results more stable and cuts down on false alarms, while still finding as many real effects as older methods.
Integrity screen: passed (3 checks) Checked 9 November 2026. Read the source
The plain-language reading of this paper is in the full issue, for subscribers.
Read the full issue