Motif relation analysis based on ICA technology

As the transcription factor binding sites,the motifs play an important role in promoter recognition and the gene transcription and expression.Finding their characteristics is obviously a meaningful subject.According on the correlation detection of motifs as promoter features,a method for getting motif packages is given by independent component analysis.It can decompose the obverved motif frequency matrix into smaller components and finally obtain the motif packages with concurrent motifs.In the experiment results,the influence of DNA sequence data selection to the motif packages is analyzed.More stable motif relation structures can be obtained using longer motifs.