An integrated in silico approach to analyze the involvement of single amino acid polymorphisms in FANCD1/BRCA2-PALB2 and FANCD1/BRCA2-RAD51 complex.

Fanconi anemia (FA) is an autosomal recessive human disease characterized by genomic instability and a marked increase in cancer risk. The importance of FANCD1 gene is manifested by the fact that deleterious amino acid substitutions were found to confer susceptibility to hereditary breast and ovarian cancers. Attaining experimental knowledge about the possible disease-associated substitutions is laborious and time consuming. The recent introduction of genome variation analyzing in silico tools have the capability to identify the deleterious variants in an efficient manner. In this study, we conducted in silico variation analysis of deleterious non-synonymous SNPs at both functional and structural level in the breast cancer and FA susceptibility gene BRCA2/FANCD1. To identify and characterize deleterious mutations in this study, five in silico tools based on two different prediction methods namely pathogenicity prediction (SIFT, PolyPhen, and PANTHER), and protein stability prediction (I-Mutant 2.0 and MuStab) were analyzed. Based on the deleterious scores that overlap in these in silico approaches, and the availability of three-dimensional structures, structure analysis was carried out with the major mutations that occurred in the native protein coded by FANCD1/BRCA2 gene. In this work, we report the results of the first molecular dynamics (MD) simulation study performed to analyze the structural level changes in time scale level with respect to the native and mutated protein complexes (G25R, W31C, W31R in FANCD1/BRCA2-PALB2, and F1524V, V1532F in FANCD1/BRCA2-RAD51). Analysis of the MD trajectories indicated that predicted deleterious variants alter the structural behavior of BRCA2-PALB2 and BRCA2-RAD51 protein complexes. In addition, statistical analysis was employed to test the significance of these in silico tool predictions. Based on these predictions, we conclude that the identification of disease-related SNPs by in silico methods, in combination with MD approach has the potential to create personalized tools for the diagnosis, prognosis, and treatment of diseases. The methods reviewed here generated a considerable amount of valuable data, but also the need for further validation.