Prediction the functional impacts of highly deleterious non-synonymous variants of TSGA10 gene.

Abstract

Testis specific gene antigen 10 (TSGA10) is a protein which has roles in spermatogenesis and cancers so that deletion or mutation in the TSGA10 gene resulted in non-obstructive infertility and aberrant expression of this protein, was detected in solid tumors and leukemia. Despite the crucial roles of TSGA10 in tumorigenesis and infertility, yet it is not obvious how various nsSNPs of its gene impress the structure and function of the TSGA10. Therefore, it is worthwhile to investigate the potential highly deleterious nsSNPs by several in-silico tools before launching costly experimental approaches. In the current study, we employed several different machine learning algorithms in a two-step screening procedure to analyze single nucleotide substitutions of TSGA10 gene. Prediction tools were included SIFT, PROVEAN, PolyPhen-2, SNAP2, SNPs & GO, PhD-SNP for the first step and the second step included predictive tools such as I-mutant 3.0, MUpro, SNPeffect 4.0 (LIMBO, WALTZ, TANGO, FoldX), MutationTaster and CADD. Also, the 3D models of significantly damaging variants were built by Phyre2. The results elucidated 15 amino acid alterations as the most deleterious ones. Among these S563P, E578K, Q580P, R638L, R638C, R638G, R638S, L648R, R649C, R649H were located in a domain which is approved to has interaction with the HIF1-A protein and D62Y, R105G, D106V and D111Y were located on phosphodiesterase domain. In sum, these predicted mutations significantly influence the function of TSGA10 and they could be used for precise study of this protein in infertility and cancer experimental investigations.