Fig. 2: Variant prioritization and burden testing in NOA cases.

a Distribution of the prioritized variation types across inheritance modes. AR autosomal recessive, Chet Compound heterozygous, XL X-linked, YL Y-linked. b Predicted hemizygous loss-of-function genotypes appear to be enriched on sex chromosomes compared to biallelic loss-of-function genotypes on autosomes (A) (Fisher’s exact p = 0.018 for chrX and p = 0.032 for chrY vs A). *p < 0.05. c Summary of all genes with multiple case findings in the GEMINI cohort. d PSAP p values of all variants detected in patient GEMINI-295, the carrier of the biallelic STRA8 deletion. STRA8 CNV (the smallest p value) was prioritized as the most likely cause of NOA in GEMINI-295. The gray area represents the 95% confidence interval. e Z-scores of normalized read depth of exome sequencing data around the STRA8 locus, plotted against the STRA8 gene model. NLS nuclear localization; NES, nuclear export signal. Green arrows, PCR primers. f PCR spanning the predicted deletion region yielded a short ~5 kb product in GEMINI-295 (P) validating the homozygous STRA8 deletion. The experiment repeated twice with the same result. C, control; M, ladder. g In control testis, clear staining of STRA8 was observed in spermatogonia (arrows), together with background staining in peritubular and interstitial cells (asterisk). In GEMINI-295, STRA8 protein staining in spermatogonia is much lower (arrows), presumably due to nonsense-mediated decay of the transcript. Three different sections of the controls (n = 8) and the case (n = 1) showed consistent staining patterns. The bar represents 10 µm. h Burden test results from comparison of a combined cohort of 2072 NOA cases with 11,587 fertile controls. Thirty-four genes with prioritized variation in the GEMINI cohort were selected for burden testing; of these 21 were nominally associated with NOA (“Methods”). * indicates genes with at least “limited” clinical validity defined by ref. 7.