Membrane protein classification by machine learning techniques

Abstract

One of the indispensable application of computer science is in the and#64257;eld of biology, newlinewhich has given rise to the new science - Bioinformatics. It might be appropriate to newlinesay that Bioinformaticians deals with application of computers in solving problem newlineof molecular biology. Bioinformatics are mainly concerned with problems involving newlinedata emerging from within cells of living beings. The data emerging from a cell are newlinebasically of four types : DNA, RNA, Protein sequences and Micro array images. newlineHowever, the and#64257;rst three of them are mere text data and fourth one is image data. newlineProteins are important molecules in life. They are the work horse in any living newlineorganism. Our body is just a protein factory, capable of producing 100,000 vivid newlineproteins. When they are produced in the right time, at right place, in the right newlinequantity, we are healthy. The vast majority of the targets of approved drugs are newlineproteins. So proteins play a very important role in drug development. newlineProteins are made up of amino acids, which are twenty in count. Proteins are not newlinelinear chains of amino acids. They are famous for their shapes. They turn, twist newlineand fold into very unique shapes. These shapes are studied at 4 levels primary, newlinesecondary, tertiary and quaternary. newlineProteins are classiand#64257;ed into diand#64256;erent categories based on their structure ,compo newlinesition and function. Each category are divided into diand#64256;erent classes. One such class newlinebased on structure classiand#64257;cation category is membrane protein, which has further newlinesubdivisions(subclasses). A membrane protein can belong to more than one subclass newlinesimultaneously. Hence classiand#64257;cation of membrane proteins in to diand#64256;erent subclasses newlinebelong to multi label classiand#64257;cation problems. newline newline