MAnorm_2

MAnorm_2 performs quantitative differential analysis of groups of ChIP-seq samples to identify differential binding sites of chromatin-associated proteins across cellular contexts.

Key Features:

• Hierarchical Normalization Strategy: Implements a hierarchical normalization to normalize ChIP-seq signal across groups and account for variability among biological replicates and experimental conditions.
• Empirical Bayes Framework: Uses an empirical Bayes framework to assess within-group variability of ChIP-seq signals and enhance detection of true differential binding events.
• Handling Biological Replicates: Supports analysis of biological replicates to detect differential binding even between highly similar cellular contexts and subtle changes in protein–DNA interactions.
• Performance with Variable Within-Group Variability: Exhibits improved differential-detection performance relative to existing methods, particularly when comparing groups with distinct global within-group variability.

Scientific Applications:

• Inference of Differential Binding Sites: Inferring differential binding sites between cellular contexts using ChIP-seq datasets.
• Study of Transcriptional Regulation: Investigating regulation of eukaryotic gene transcription by chromatin-associated proteins through changes in binding patterns.
• Detection of Subtle Protein–DNA Interaction Changes: Detecting subtle changes in protein–DNA interactions across biological replicates and closely related cell types.

Methodology:

Preprocessing converts ChIP-seq samples into a structured format that records read abundances and enrichment states across genomic bins via the MAnorm2_utils package used alongside the main R package, followed by hierarchical normalization and empirical Bayes-based differential analysis.