DNA Sequence Motif Discovery Using Greedy Construction Algorithm Based Techniques

DNA motif discovery is an important problem in bioinformatics and it is essential for identifying transcription factor binding sites that play key role in the gene expression process. Motifs are generally short patterns which repeat among a set of DNA sequences. However, it is computationally expensive and impractical to find them with exhaustive search. Therefore, probabilistic and heuristic approaches can be used for tackling this problem. This study focuses on greedy construction algorithms for finding DNA motifs. First, we develop a GRASP algorithm with candidate list reduction with a one-sequence- change neighborhood based hill climbing local search. Then, we evaluate and discuss the performance of the developed algorithm on 3 common datasets, namely hm03r, yst04r, and yst08r. The results obtained show the effectiveness and usability of the proposed method for the DNA motif discovery problem.