diagFDR
Provides scope, calibration and stability diagnostics to assess false discovery rate (FDR) claims in peptide and protein identification workflows based on target-decoy competition and related confidence measures. Includes the recomputed decoy ratio of the reported list, accepted decoy counts and local decoy support, the FDP-CV benchmark, threshold elasticity, posterior error probability (PEP) concordance with the decoys, equal-chance checks and p-value calibration, explicit target/decoy label-propagation rules for grouped units, picked-protein competition, external validation by entrapment (raw and size-corrected estimates, scope-discordant-set error rate), a ground-truth simulator of target-decoy competition, and an HTML report. Methods are described in Chion et al. (2026) <doi:10.64898/2026.04.16.718468>.
This page was built for software: diagFDR
