Based On The Motif Of The Gibbs Sampling Algorithm To Find New Methods Of Research

One important problem in bioinformatics is to understand how genes cooperate to perform functions, i.e. the gene regulatory network. Since the experimental determination of motif is expensive and time-consuming, the computational methods are developed greatly. Usually, the computational methods assume that through long time evolution, the gene which is regulated by the co-regulated gene clusters contains conserved gene cluster called motif. Co-regulated genes contains motif in the upstream promoter element with a high possibility.Motif is a kind of short nucleotide, which have the character of conservative and maybe have some biological function. For simplifying the model of the motif, we assume that the regulate element use the simple regular model from the upstream sequence without influence by the remote control. The shared gene may have the share motif, thus we can find the motif by finding the common motif from the upstream of a series of sequences which have Co-regulated genes.In this paper, we have an overview of the situation of the popular methods to find the motif. Then we point out the disadvantages of these algorithms. According to what we found before, we developed a PWM (position weight matrix)_based algorithm to improve the Gibbs sampling algorithm for the motif discovery. In this algorithm, we introduce the motif base to record the statistic data. And we try to verify the effective of the algorithm with the motif which belongs to the database verified by the biology test. In the results, we have found some advantages in this algorithm in express the conservation and precision when compared with other algorithms.Secondly, an improved Gibbs sample algorithm is proposed to discover motif with the new method Markov chain background. A new method is been proposed to encode the candidate motif. The Markov factor is adopted to reduce the error which has a great influence on the result of the computer. The verified biological data are used to test the feasibility and effectiveness of improved approach. Compared with results given by the experiment, our algorithm is proved to raise the accuracy, flexibility and stability for the motif discovery effectively.