Table 1.

OddsPath-based assay evidence calibration using RAD51D clinical variants

Pathogenic variants					Benign variants
n	Abnormal	Normal	Odds Path	Evidence points	n	Abnormal	Normal	Odds Path	Evidence points
37	37	0	174	+4	174	0	174	1/37	−4

Known pathogenic or benign variants were obtained from ClinVar, requiring two stars (multiple submitters, without discordance), and resolving conflicts when at least two non-VUS interpretations were available (Methods). OddsPath scores were determined as described²⁶. Variants predicted to be splice disruptive by SpliceAI (score≥0.2) were excluded and treated as ‘not determined’.