A novel gene identification algorithm with Bayesian classification

The rapid advances in the field of computational genomics and bioinformatics have motivated the development of innovative engineering methods for data acquisition, interpretation, and analysis. With the help of the later methods, many processes in molecular biology can be modeled and further analyzed. Identification and discovery of the coding regions in the genomic structure using computational algorithms is a clear example of such processes. This work proposes a novel application of well-known principles and concepts from communications theory and digital signal processing for the detection of protein coding regions in prokaryotic genomes. The proposed algorithm employs a polyphase complex mapping scheme to provide a numerical representation of the genomic sequences involved in the analysis. It then utilizes concepts in communications theory such as correlation, the maximal ratio combining (MRC) algorithm, and filtering to generate a signal whose peaks and troughs signify coding and noncoding regions, respectively. The proposed algorithm is applied to several prokaryotic genome sequences. Two Bayesian classifiers are designed to evaluate the performance of the proposed algorithm. The obtained simulation results show that the algorithm is able to efficiently and accurately identify protein coding regions with sensitivity and specificity values comparable to well-known gene detection methods in prokaryotes such as GLIMMER and GeneMark. This further proves the relevance of using communications theory concepts for genomic sequence analysis.