Effectiveness of Applying Codon Usage Bias for Translational Initiation Sites Prediction

The accurate recognition of translational initiation sites (TISs) in genomic, cDNA and mRNA sequences is crucial to identifying the primary structure of the functional gene products - proteins. Many computational methods have been proposed in the literature which apply one or more complicated models that examine a variety of sequence features. In this paper, we propose a novel TIS prediction approach, called codon usage bias agent, which operates solely based on the usage of codon preference; the algorithm only requires O(n) for execution time. The results of the experiments conducted on three benchmark data sets have shown that the proposed approach is very effective and well applicable to solving the problem of TIS recognition.