Abstract—Dramatic advances in genomics and computational biology have resulted in large amounts of data and have encouraged the development of computational algorithms for the identification and analysis of coding regions. This paper proposes a novel application of fundamental principles and concepts from communications theory for the identification of exact translation initiation sites in prokaryotic genomes. It employs several Bayesian classifiers to assess the performance of the ribosome binding sites detection algorithms investigated in this work. The proposed classification algorithms utilize well-known principles in communications theory such as cross correlation and Euclidean distance based metrics to make precise real-time decisions of weather a given open reading frame (ORF) is a valid protein coding region or not. The simulation results confirm that the proposed Bayesian classification algorithms can provide a efficient and accurate gene identification with sensitivity and specificity values comparable to the ones obtained by the well-known prokaryotic gene detection methods such as GLIMMER and GeneMark. This further confirms the significance of applying communications theory concepts to genomic sequence analysis.

Index Terms—Gene detection, cross correlation, Euclidean distance, Bayesian classification.

Mohammad F. Al Bataineh is with Telecommunications Engineering Department, Yarmouk University, Irbid, Jordan (email: mohamadfa@yu.edu.jo).
Zouhair J. Al-Qudah is with Department of Communication Engineering, Al-Hussein bin Talal University, Ma'an, Jordan.

Cite: Mohammad F. Al Bataineh, Zouhair J. Al-Qudah, "Bayesian Classification of Ribosome Binding Sites in Prokaryotic Genome Sequences: A Communications Theory Approach," International Journal of Bioscience, Biochemistry and Bioinformatics vol. 7, no. 3, pp. 133-142, 2017.