SAVDERMA: AN INTEGRATED TOOL AND DATABASE FOR SINGLE AMINO ACID VARIATIONS ASSOCIATED WITH DERMATOLOGICAL DISORDERS

Abstract

Single Amino Acid Variations (SAVs) also called non-synonymous single nucleotide polymorphisms SNPs (nsSNPs), are the most inexhaustible type of single nucleotide polymorphisms (SNPs) that lead to amino acid substitutions in the protein products. Among different SAVs, some may cause pernicious infections and diseases while other amino acid substitutions are neutral which won't influence the capacity of the protein. Various researches on protein structures and functions have proposed that some SAVs are accountable for certain diseases, and it is responsible for about 60% of Mendelian diseases caused by amino acid substitutions. Studies on SAVs also showed that a mutation in an amino acid sequence leads to different skin disease. SAVs are of utmost relevance in checking their association with disease. SAVDerma is an integrated tool and database which contains information about SAVs and their association with Dermatological disorders. Data related to SAVs were retrieved through various databases and used to prepare training and testing data for machine learning. Various classifiers used to train data, among them Random forest classifier is used for testing data because of its highest accuracy of 87.29 percent and highest precision of 87.40 percent in predicting the association of SAVs in dermatological disorders. After compiling information on SAVs through various databases and utilising prediction data which gives insight about an SAV and its association with dermatological disorders, SAVDerma was developed. SAVDerma compiles highly curated physicochemical and sequence-based data of SAVs. Users can retrieve information regarding a particular SAV through three different parameters (HumsaVar FT id, Uniprot id and Swissport id). Results will show amino acid position at which variation is present along with their association with dermatological disease.