Finding genomic features from enriched regions in ChlP-Seq data

Finding genomic features in ChlP-Seq data has become an attractive research topic lately, because of the power, resolution and low-noise of next generation sequencing, making it a much better alternative to traditional microarrays such as ChlP-chip and other related methods. However, handling ChlP-Seq data is not straightforward, mainly because of the large amounts of data produced by next generation sequencing. ChlP-Seq has widespread over a range of applications in finding biomarkers, especially those associated with important genomic features in epigenomics and transcriptomics, including binding sites, promoters, exons/introns, transcription sites, among others. Efficient algorithms for finding relevant regions in ChlP-Seq data have been proposed, which capture the most significant peaks from the sequence reads. Among these, multilevel thresholding algorithms have been applied successfully for transcriptomics and genomics data analysis, in particular for detecting significant regions based on next generation sequencing data. We show that the Optimal Multilevel Thresholding algorithm (OMT) achieves higher accuracy in detecting enriched regions and genomic features of detected regions on FoxAl data. OMT finds more gene-related regions (gene, exon, promoter) in comparison with other methods. Using a small number of parameters is another advantage of the proposed method.