Binary classification of protein sequence using machine learning

Abstract

Bioinformatics work with so many technical domains. DNA, Gene datasets al ready done prediction model using machine learning techniques. Protein sequences binary classi cation using machine learning, now a days its use for to predict accu racy or e ciency of proteins for particular protein family, human body have thou sands of protein families and inside that crore of proteins are there so, for the analysis of some disease and analysis of proteins we have nd a way like classi cation prob lem. For example there are to many individual web portals available which gives accuracy of protein sequences. In binary classi cation of protein sequence using machine learning algorithm gives an information of proteins which are belong from which protein family or not, when we nd classifying accuracy using algorithms from that we can get information of correctly and incorrectly classi ed instances. Protein sequence classi cation generally based on their features, its a biggest task over here which and what kind of features we use and from that we can get higher accuracy results. Here, we extracts protein features are amino acid, dipeptide, pseudo dipeptide and Position speci c scoring matrix compositions found and then gives this vectors data as input of machine learning techniques. Binary classi cation done by so many algorithms among all the algorithms we chose four algorithms for analysis purpose. Support Vector Machine(SVM), Naive Bayes classi er, Logistic regression and Linear Discriminant Analysis(LDA) use as a classi er for binary classi cation