consensusSeekeR

consensusSeekeR identifies consensus regions by analyzing genomic positions and ranges across multiple experimental datasets to detect common regions such as nucleosome positions and chromatin features.

Key Features:

• Consensus region detection: Analyzes genomic positions and ranges across multiple experiments to identify common regions (consensus regions).
• Adjustable parameters: Allows adjustment of analyzed region size and the number of experiments required for a feature to be considered part of a consensus region.
• Bayesian nucleosome positioning: Employs a Bayesian method to pinpoint nucleosome positions.
• Statistical modeling: Uses Multinomial-Dirichlet classification combined with hierarchical mixture distributions for modeling signal and feature classes.
• Model selection via MCMC: Estimates both the number and precise locations of nucleosomes using reversible jump Markov chain Monte Carlo simulation.
• Compatibility with mapping technologies: Applicable to genome-wide mapping data including ChIP-chip, ChIP-Seq, and MNase-Seq.
• Benchmarking: Validated on simulated data and MNase-Seq from Saccharomyces cerevisiae and compared against methods such as PING and NOrMAL.

Scientific Applications:

• Nucleosome mapping: Identification and precise localization of nucleosome positions from MNase-Seq and related datasets.
• Consensus discovery across experiments: Detection of shared chromatin features across multiple ChIP-chip or ChIP-Seq experiments.
• Chromatin organization studies: Comparative analysis of chromatin structure to investigate relationships with gene regulation and transcription factor access.
• Method benchmarking: Quantitative performance comparison against existing methods such as PING and NOrMAL using simulated and experimental data.

Methodology:

Applies a Bayesian framework using Multinomial-Dirichlet classification and hierarchical mixture distributions, with reversible jump Markov chain Monte Carlo to estimate the number and locations of nucleosomes.