Fig. 4

Identification of pathogenic mutations in TUBA4A. a Estimation of the damaging recurrent DNM genes. P-values and q values for recurrence were calculated using TADA (the de novo-only algorithm) in our cohort. Based on previously published thresholds (q ≤ 0.1), PCDH20 (q = 0.42) and DGKZ (q = 0.51) did not meet this threshold and were removed from downstream analyses. b Enrichment of four candidate genes compared with seven external de novo datasets, including male infertility (MI), undiagnosed developmental disorders (UDD), autism spectrum disorder (ASD), schizophrenia (SCZ), congenital heart disease (CHD), Tourette disorder (TD), and unaffected controls (CT). x axis was labeled as “disease_sex_sample number” and mix means mixed sample. P-values were calculated using the Poisson cumulative density function in external datasets. c Three infertile parent-child trios with de novo TUBA4A mutations. Squares denote male family members, circles denote female members, and solid symbols represent affected members. The “ = ” sign indicates infertility. d Nine pedigrees of sporadic TUBA4A mutations. Squares denote male family members, circles denote female members, and solid symbols represent affected members. The “ = ” sign indicates infertility