HMCan

HMCan detects histone modification signals in cancer-derived ChIP-seq data by correcting copy number, GC-content, and noise biases to enable accurate identification of epigenetic marks.

Key Features:

• Correction for Biases: Implements a three-step correction process (copy number, GC-content, and noise-level corrections) to improve histone modification detection accuracy in cancer genomes.
• Copy Number Correction: Adjusts for chromosomal gains and losses that can skew ChIP-seq signal enrichment evaluations, preventing under- or over-detection of histone marks.
• GC Bias Correction: Accounts for GC-content variation across genomic regions that affects ChIP-seq read distribution and signal interpretation.
• Noise Level Correction: Reduces background noise to improve the clarity and reliability of detected signals.
• Application of Hidden Markov Models: Employs Hidden Markov Models to segment the corrected signal and identify histone modification regions in cancer genomes.
• Superior Performance: Demonstrated superior performance on simulated datasets and real ChIP-seq data, including the H3K27me3 mark in a bladder cancer cell line, with better alignment to qPCR-validated regions.
• Relevance to Cancer Research: Enables identification of epigenetic changes such as local or regional silencing of tumor suppressor genes.

Scientific Applications:

• Oncogenesis studies: Detecting histone modifications that may contribute to oncogenesis.
• Tumor suppressor gene analysis: Identifying regions with aberrant epigenetic silencing of tumor suppressor genes.
• ChIP-seq analysis in cancer samples: Providing more accurate analysis of ChIP-seq data from cancer samples complicated by copy number alterations and other genomic changes.

Methodology:

Applies three computational corrections (copy number, GC-content, and noise-level) to ChIP-seq signal followed by Hidden Markov Model–based detection of histone modification regions.