hmm_bigwigs

A Python tool to perform Hidden Markov Model (HMM) analysis on genomic data stored in bigWig files. This tool uses the pomegranate library to identify genomic states based on signal intensity patterns.

Installation

Install using pip:

pip install hmm_bigwigs

Requirements

• numpy
• pandas
• bioframe
• pomegranate
• pybbi

Usage

bigwig_hmm.py -i <input.bw> -g <genome> -n <num_states> -o <output_prefix>

Arguments

• -i, --inputfile (required): Path to input bigWig file
• -g, --genome: Reference genome (e.g., hg38, mm10)
• -n, --num_states: Number of HMM states to identify (default: 2)
• -o, --outputfile: Output file prefix (will generate files ending in "_state_HMM.bed")
• --cmap: Matplotlib colormap name for state visualization (default: 'coolwarm')

Output Files

The tool generates several BED format files:

• {prefix}_n_state_HMM_colored.bed: Complete segmentation with all states
• {prefix}_0_state.bed: Regions assigned to state 0
• {prefix}_1_state.bed: Regions assigned to state 1
• {prefix}_2_state.bed: Regions assigned to state 2 (if n_states > 2)

Each BED file contains the following columns:

• chromosome
• start position
• end position
• state assignment
• score (0)
• strand (.)
• RGB color values

Example

bigwig_hmm.py -i H3K27me3.bw -g hg38 -n 3 -o H3K27me3_domains

This command will:

1. Read the H3K27me3.bw bigWig file
2. Segment the genome into 3 states
3. Generate color-coded BED files for visualization

Features

• Automatic handling of chromosome selection (excludes chrM, chrX, chrY by default)
• Intelligent state assignment based on signal intensity
• Color-coded output for easy visualization in genome browsers
• Efficient merging of adjacent regions with identical states

License

[Add your license information here]

Citation

If you use this tool in your research, please cite: [Add citation information here]

Contact

• Author: George Spracklin
• Issues: https://github.com/gspracklin/hmm_bigwigs/issues