Bidirectional filtering approach for the improved protein coding region identification in eukaryotes

AbstractThe Gene identification problem or the computational techniques for finding the protein-coding regions in a given DNA sequence is one of the extensive research issues in bioinformatics. Conventional statistical methods that are used to identify the protein-coding region from DNA sequences are expensive and time-consuming. Over the past two decades, there has been an increasing interest in exploring DSP-based solutions to this problem. A class of infinite impulse response (IIR) anti-notch filters (ANF) has been previously used to meet the requirements. However, the prediction accuracy of these filters is still limited due to the inherent nonlinear phase delay and distortion. This paper seeks the remedy of this problem by using a bidirectional filtering approach. The obtained results support the idea of bidirectional filtering as the prediction performance of conventional ANF is improved by 12 –17%.
Source: Network Modeling Analysis in Health Informatics and Bioinformatics - Category: Bioinformatics Source Type: research