Peak Calling - Usage Guide
Overview
Call peaks from ChIP-seq data using MACS3 to identify transcription factor binding sites or histone modification regions.
Prerequisites
# Conda (recommended)
conda install -c bioconda macs3
# Pip
pip install macs3
Quick Start
Tell your AI agent what you want to do:
- "Call peaks from my ChIP-seq BAM file"
- "Identify H3K27ac enriched regions"
- "Find transcription factor binding sites with input control"
Example Prompts
Narrow Peaks (TFs, H3K4me3)
"Call narrow peaks for my transcription factor ChIP-seq"
"Run MACS3 on my H3K27ac ChIP-seq with input control"
Broad Peaks (H3K27me3, H3K36me3)
"Call broad peaks for my H3K27me3 ChIP-seq data"
"Identify H3K9me3 domains using broad peak calling"
ATAC-seq
"Call peaks from my ATAC-seq data"
Troubleshooting
"My peak calling found no peaks, help me troubleshoot"
"I'm getting too many peaks, how can I increase stringency?"
What the Agent Will Do
- Sort and index BAM files if not already done
- Select appropriate peak type (narrow vs broad) based on target
- Run MACS3 with appropriate parameters and genome size
- Generate narrowPeak/broadPeak files and summit BED files
- Report peak counts and quality metrics
- Suggest parameter adjustments if results are unexpected
Tips
- Always use input/IgG control for TF ChIP-seq
- Use
--broadfor histone marks that form domains (H3K27me3, H3K36me3) - Default q-value threshold is 0.05; use
-q 0.01for higher stringency - Use
--nomodel --extsize 200if model building fails on small datasets - Check peak numbers: TFs typically have 1,000-50,000 peaks; broad marks can have fewer but larger regions