ChIPmeta

ChIPmeta integrates ChIP-seq and ChIP-chip data to improve identification of transcription factor binding sites (TFBS) by modeling technology-specific signals.

Key Features:

• Data integration: Combines ChIP-seq (chromatin immunoprecipitation followed by sequencing) and ChIP-chip (chromatin immunoprecipitation followed by array hybridization) datasets to leverage complementary strengths and address technology-specific limitations.
• Sequencing and array characteristics: Accounts for ChIP-seq high-resolution genome-wide coverage down to single base pair precision and ChIP-chip array-based hybridization signals.
• Hierarchical Hidden Markov Model (HHMM): Uses a higher-level HHMM to summarize and integrate inference results from individual hidden Markov models (HMMs) applied to each data type.
• Modeling of technology-specific signals: Captures nuanced differences between ChIP-seq and ChIP-chip signals to enhance TFBS detection.
• Evaluation and validation: Includes simulation studies demonstrating improved performance when both data types are available and empirical analyses showing enhanced TFBS identification.

Scientific Applications:

• Transcription factor binding site identification: Improves detection of TFBS for transcription factors such as NRSF (Neuron-Restrictive Silencer Factor) and CTCF (CCCTC-binding factor) by integrating multiple experimental platforms.
• Gene regulation studies: Supports genomic analyses that require comprehensive and accurate mapping of TFBS to investigate gene regulatory mechanisms.

Methodology:

Individual HMMs are applied to ChIP-seq and ChIP-chip data and their inference results are summarized by a higher-level hierarchical hidden Markov model (HHMM); performance was assessed by simulation studies and empirical analyses on NRSF and CTCF.