deeptools

deepTools: NGS sequencing data analysis and visualization tools

Skill overview

deepTools is a suite of Python command-line utilities designed for high-throughput sequencing data, providing quality control, normalization, sample comparison, and publication-ready visualization for experiments such as ChIP-seq, RNA-seq, and ATAC-seq.

Applicable scenarios

1. ChIP-seq data analysis workflow

One-stop processing from BAM files to signal visualization. Supports sample correlation analysis, PCA dimensionality reduction, enrichment intensity assessment, and generation of heatmaps and signal profiles over regions like TSS and peaks to quickly validate ChIP experiment quality.

2. RNA-seq coverage and visualization

Supports strand-specific coverage calculation to distinguish sense and antisense signals, generates coverage tracks over gene bodies and exon regions, suitable for visualizing transcriptome expression patterns.

3. ATAC-seq chromatin accessibility analysis

Includes built-in Tn5 transposase offset correction to accurately analyze open chromatin regions. Supports fragment length distribution analysis (nucleosome ladder patterns), heatmaps of accessibility sites, and coverage visualization.

Core features

BAM/bigWig file processing

Convert BAM alignments to normalized coverage tracks (bigWig/bedGraph formats), supporting normalization methods such as RPGC, CPM, and RPKM. Can generate log2 ratio tracks between samples, perform GC-bias correction, and provide multi-sample summary statistics.

Quality control and sample comparison

Provides plotFingerprint to assess ChIP enrichment, plotCorrelation to analyze sample correlations, plotPCA for principal component analysis, plotCoverage to check sequencing depth, and bamPEFragmentSize to validate fragment length distribution. Related samples should show high correlation (>0.9), and strong ChIP signal should display a steep enrichment curve.

Visualization and enrichment analysis

Use computeMatrix to calculate signal matrices over genomic features, plotHeatmap to generate clustered heatmaps, plotProfile to draw signal distribution curves, and plotEnrichment to analyze enrichment in peak regions. Supports K-means clustering, custom color maps, and multi-sample overlay displays.

Frequently asked questions

What sequencing data types does deepTools support?

It supports various high-throughput sequencing data including ChIP-seq, RNA-seq, ATAC-seq, MNase-seq, and Hi-C. Input formats are BAM alignment files and BED/GTF region annotation files; outputs include bigWig tracks, PNG/SVG images, and statistical results.

How to choose the correct normalization method?

For ChIP-seq coverage, RPGC (1x genome coverage) or CPM are recommended; for RNA-seq bins use CPM, and for gene-level expression use RPKM (accounting for gene length); for ATAC-seq use RPGC or CPM. For sample comparisons, using bamCompare with log2 mode together with readCount scaling factors is recommended.

What is the difference between deepTools and IGV?

deepTools is a command-line batch processing toolkit suited for multi-sample QC statistics and bulk generation of standardized plots; IGV is an interactive browser suited for detailed single-sample inspection and exploratory analysis. They complement each other—deepTools quickly screens problematic samples, and IGV is used for in-depth validation of details.