Functional consequence analysis of nsSNPs

Sort Intolerant From Tolerant (SIFT) (https://sift.bii.a-star.edu.sg/) was employed to detect the deleteriousness of nsSNPs. SIFT can distinguish the deleterious and neutral effects of amino acid substitutions in nsSNPs and missense mutations based on physical characteristics and sequence homology of amino acids [23]. It utilizes multiple sequence alignment to obtain normalized probability scores for all substitutions. A score <0.05 is considered a deleterious substitution.

Polymorphism Phenotyping v2 (PolyPhen-2) (http://genetics.bwh.harvard.edu/pph2/) is a publicly accessible web server for predicting the structural and functional consequences of amino acid substitutions [24]. Variants with PolyPhen-2 score of (0.0–0.15) are considered benign, (0.15–1.0) as possibly damaging, and (0.85–1.0) as damaging.

The Rare Exome Variant Ensemble Learner (REVEL) (https://sites.google.com/site/revelgenomics/) is an ensemble method for detecting the pathogenic nsSNPs based on tools, namely MutPred, PolyPhen, FATHMM, SIFT, MutationAssessor, PROVEAN, and several ensemble methods. REVEL score ranges from (0–1) with a cut-off of 0.5 [25].

MetaLR (https://wglab.org/) distinguishes between neutral and damaging SNPs using logistic regression by providing a score between 0 to 1, where a score>0.5 indicates the damaging effect [26]. MutationAssessor (http://mutationassessor.org/r3/) is a web server that estimates the functional effect of missense polymorphisms and mutations based on evolutionary conservation in protein homologs. It produces a score ranging from 0 to 1. nsSNPs with higher scores are more likely to be pathogenic [27].

MutPred2 (http://mutpred.mutdb.org/), a machine learning-based method, estimates the pathogenicity and molecular alteration of single nucleotide polymorphisms by integrating genetic and molecular data [28]. MutPred2 generates a general score from the mean scores of the neural networks. A score cut-off of 0.50 denotes pathogenicity. Protein ANalysis THrough Evolutionary Relationships (PANTHER) (http://www.pantherdb.org/tools/) is a comprehensive, freely available database that employs phylogenetics to analyze protein sequences and determine their evolutionary links to other proteins [29]. It employs PANTHER-PSEP (Position-Specific Evolutionary Preservation) to anticipate how nonsynonymous coding single nucleotide polymorphisms may affect the functionality of proteins [30].

ClinVar (https://www.ncbi.nlm.nih.gov/clinvar/) is a public database of genetic variants and their clinical significance that gathers data from a variety of sources, such as clinical testing facilities, research projects, and the scientific literature, and disseminates knowledge regarding the associations between genetic variants and diseases or other health issues [31]. PON-P2 (http://structure.bmc.lu.se/PON-P2/) is a machine learning-based tool that has been developed for the classification of amino acid substitutions in human proteins, utilizing the evolutionary conservation of sequences, the physical and biochemical properties of amino acids, Gene Ontology (GO) annotations, and functional annotations of variation sites [32].