Klasifikasi Protein Family Menggunakan Metode Rantai Markov

Abstract

Proteins are large biological molecules that consist of one or more chains of 20 amino acids sequences. Proteins have various functions such as catalyzing metabolic reactions, replicating DNA, and transporting molecules. Many new protein sequences have been found. Therefore, a new method is required to classify proteins which have similar amino acids sequence, structure and function into the same class. The aim of this research is to apply Markov chain concept for protein family classification based on amino acid sequences. Protein data samples are obtained from Pfam protein family database website. Samples are taken from three different protein family classes, namely: 1-cysPrx_C, 4HBT and ABC_Tran, 100 data samples each. This research utilizes two different extraction methods: first order and second order Markov chain. First order indicates that the data sequence stored in the transition matrix is every two adjacent sequences, while the second order indicates that the data sequence stored in the transition matrix is every two sequences that are intermittent twice. The results show that the first order method gives 96% as its best accuracy, whereas the second order method gives 89.33% as its best accuracy