An Improved Graph Algorithm In Predicting Regulatory Elements

With the rapid development of bioinformatics, revealing the regulate mechanism of gene expression becomes a main challenge in molecular biology. One of the most important steps is to find the regulators, especially the binding sites of the transcription factor. Transcription factors are the protein normally binding the upstream DNA sequences, especially near the transcription start site. They regulate the expression of protein by regulating the transcription mechanism or inhibiting transcription. It is still difficult to predict these regulatory elements. Even the most thorough studied organisms, we still know little about their transcription factor and its binding sites. To accurately identify these sites is very difficult because they are generally very short, only about 10 bases, but generally the input sequence length is 1000bp. What makes the problem more complicated is that each transcription factor has a variety of binding sites, and there are changes in the sequences of binding sites. In this study, with the weighted graph theory, using the improved scoring function to create a new algorithm, we scored each two fragments of the input sequences,We used 5 transcription factors of Nematode to test this algorithm, and selected the other three algorithms to compare, AlignACE, Consensus, and Gibbs respectively. This algorithm has shown a higher accuracy and sensitivity. As for different species, we can adjust the data used in this method to make it applicable to other species, so that the algorithm is applicable to a wider range。...