epigenomix

epigenomix integrates RNA sequencing (RNA-seq) or microarray transcription data with Chromatin Immunoprecipitation Sequencing (ChIP-seq) histone modification data to identify genes whose transcriptional changes are associated with histone modification patterns.

Key Features:

• Data Integration: Preprocesses and aligns RNA-seq or microarray data with ChIP-seq data to enable joint analysis of transcription levels and histone modification patterns.
• ChIP-seq Normalization: Implements normalization strategies for ChIP-seq datasets to ensure accurate downstream correlation measures.
• Correlation Measure: Computes a novel correlation measure specifically tailored to assess relationships between ChIP-seq signals and gene transcription across conditions.
• Bayesian Mixture Models: Uses Bayesian mixture models with multiple distribution types to model the distribution of correlation measures and implicitly classify genes exhibiting joint changes in transcription and histone modifications between two experimental conditions.
• Comparative Performance: Demonstrates improved detection of genes with coordinated transcriptional and histone modification changes compared with separate analyses of each data type.

Scientific Applications:

• Epigenetic regulation analysis: Identifying genes whose transcriptional changes are potentially regulated by histone modifications.
• Mechanistic studies: Investigating molecular mechanisms and epigenetic contributions to biological processes and diseases influenced by histone modification–mediated regulation.
• Integrative genomic studies: Combining RNA-seq, microarray, and ChIP-seq datasets to detect condition-specific coordinated changes in transcription and chromatin state.

Methodology:

Preprocessing and alignment of RNA-seq or microarray data with ChIP-seq; normalization of ChIP-seq data; computation of a novel correlation measure between ChIP-seq and transcriptional data; analysis of correlation-measure distributions using Bayesian mixture models with various distribution types and implicit classification to detect genes differing between two experimental conditions.