HPeak

HPeak identifies genomic regions enriched in Chromatin Immunoprecipitation sequencing (ChIP-seq) data to map protein–DNA interactions using a Hidden Markov Model-based peak-finding algorithm.

Key Features:

• Hidden Markov Model (HMM) peak-finding: HPeak employs an HMM-based algorithm for peak detection in ChIP-seq data.
• Model-based statistical inference: The method assumes probability distributions to perform rigorous, model-based inference of enriched genomic regions.
• Novel weighting scheme: HPeak applies a weighting scheme to sequencing read coverage to improve accuracy in inferring enrichment.
• Motif enrichment validation: Evaluations reported a higher prevalence of expected transcription factor binding motifs in HPeak-enriched sequences versus controls.
• Improved resolution versus ChIP-chip: By leveraging ChIP-seq data, HPeak achieves higher resolution along with enhanced sensitivity and specificity relative to ChIP-chip for binding site detection.

Scientific Applications:

• Protein–DNA interaction mapping: Identification of genomic binding sites for transcription factors and other chromatin-associated proteins from ChIP-seq data.
• Cistroms surveys: Directly surveying the cistroms of transcription factors and chromatin-associated proteins.
• Regulatory network analysis: Supporting studies of genetic regulation and the regulatory networks that govern gene expression.

Methodology:

HPeak applies a Hidden Markov Model for peak finding, uses a novel weighting scheme on sequencing read coverage, and conducts model-based statistical inference by assuming probability distributions to identify enriched genomic regions.