Skip to content
Get packs

Guide · Agent skill

How to use the deepTools NGS Analysis agent skill

Updated · written with AI from the project's own documentation and checked against our catalogue data

The deepTools NGS Analysis skill gives an AI agent detailed working knowledge of deepTools, a command-line toolkit for sequencing data. Once it is installed, you can ask the agent in plain language to turn alignment files (BAM) into coverage tracks (bigWig), run quality checks such as correlation, PCA and fingerprint plots, or draw heatmaps and profiles around genes or peaks. This guide covers what the skill does, how to install it, how to trigger it, and what it needs and how to use it safely.

deepTools NGS AnalysisK-Dense-AI/scientific-agent-skills · MIT

Quick start

  1. 1Download and unzip the skill. Download the ZIP from this page and unzip it into ~/.claude/skills/ or a project's .claude/skills/ folder. In the Claude apps, you can upload the ZIP under Settings → Capabilities → Skills instead.
  2. 2Install deepTools. Create a uv virtual environment with Python 3.13, install deepTools with pip through uv, activate it and check that bamCoverage runs.
  3. 3Install samtools. Install samtools for sorting and indexing. Conda can install it together with deepTools using the command in the install section.
  4. 4Check your input files. Ask the agent to run the validation script on your BAM, BED and bigWig files before any analysis.
  5. 5Make a request in plain language. Describe the task, such as converting BAM to bigWig or making a TSS heatmap, and name the assay and files.
  6. 6Generate and review the workflow. Use the workflow generator to create a shell script, read it, and check paths and the effective genome size.
  7. 7Test on a small region. Run the workflow with the --region option on a small chromosome segment before running the full dataset.

What the deepTools skill makes the agent do

deepTools is a suite of Python command-line tools for processing high-throughput sequencing data. The skill teaches an agent to use it for ChIP-seq, RNA-seq, ATAC-seq, MNase-seq and similar experiments.

The skill covers four main areas: converting BAM alignments to normalized coverage tracks (bigWig or bedGraph), quality control (fingerprint, correlation, coverage), sample comparison such as correlation and PCA, and heatmap or profile plots around genomic features and peaks. It also includes two helper scripts. One validates BAM, bigWig and BED files. The other generates bash workflow templates.

  • BAM to bigWig conversion with normalization
  • QC: correlation, PCA, coverage and fingerprint plots
  • Heatmaps and profiles around TSS or peaks
  • Validation of input files before analysis

How to install the deepTools skill

Download the skill ZIP from this page. Then choose one of the two install routes below.

For a personal install, unzip the file into ~/.claude/skills/. For a single project, unzip it into the project's .claude/skills/ folder. Each skill sits in its own folder that contains SKILL.md.

In the Claude apps, you can instead upload the ZIP under Settings → Capabilities → Skills.

  • Personal: ~/.claude/skills/
  • Project only: .claude/skills/ inside the project folder
  • Claude apps: Settings → Capabilities → Skills

Install deepTools itself so the agent can run it

The skill tells the agent how to use deepTools, but deepTools must be installed on your machine for commands to run. The source recommends a uv virtual environment with Python 3.13 and a pinned deepTools release. The source also notes that the PyPI workflow was tested on macOS.

Upstream recommends conda or bioconda for full dependency resolution, especially on shared HPC systems. Conda availability varies by platform.

  • Requires Python 3.12 or newer; the skill recommends 3.13
  • Requires pysam and pyBigWig, installed with deepTools
  • Requires samtools for sorting and indexing workflows
  • uv or conda is optional for environment setup
uv venv --python 3.13 .venv-deeptools uv pip install --python .venv-deeptools/bin/python deepTools==4.0.0 source .venv-deeptools/bin/activate bamCoverage --version

How to trigger the skill with natural requests

You do not need to name commands. Describe the task in plain language and mention the data type. The skill is meant to activate for requests about file conversion, quality control, visualization, sample comparison and full analysis workflows.

If you give the agent a BAM, bigWig or BED file, say so. The skill is written for genomics files, so requests that mention those file types or assays are the clearest triggers.

  • Convert BAM to bigWig
  • Generate coverage tracks and normalize ChIP-seq data
  • Check ChIP quality or compare replicates
  • Create a heatmap around TSS or plot ChIP signal
  • Correlate samples or run PCA analysis
  • Analyze ChIP-seq data or run a complete workflow

Run the validation and workflow scripts

Before any analysis, the skill recommends validating your files. The validation script checks that local files exist and are readable, that BAM files open with their index, that BED rows are well formed and that bigWig files open. It does not confirm the assembly or whether the data are biologically suitable.

The workflow generator writes a customized shell script for standard analyses. Run --list to see the available workflows, then generate one. Make the script executable before you run it.

  • validate_files.py checks BAM/BAI/CSI, BED rows and bigWig headers
  • workflow_generator.py --list shows available templates
  • Test on small regions first using the --region option
python scripts/validate_files.py --bam sample1.bam sample2.bam --bed regions.bed python scripts/workflow_generator.py --list python scripts/workflow_generator.py chipseq_qc -o qc_workflow.sh --input-bam Input.bam --chip-bams "ChIP1.bam ChIP2.bam" chmod +x qc_workflow.sh ./qc_workflow.sh

Choose the right normalization

Normalization is the step most likely to change your results, so ask the agent to justify its choice. The skill's quick guide suggests RPGC or CPM for ChIP-seq coverage, bamCompare with log2 and readCount for ChIP comparisons, and CPM for RNA-seq bins.

For RNA-seq gene-level quantification, the skill warns that bamCoverage RPKM scales genomic bins, not genes. Use an annotation-aware gene or transcript workflow instead. RPGC needs an effective genome size that matches your reference, so verify that value before use.

  • ChIP-seq coverage: RPGC or CPM
  • ChIP-seq comparison: bamCompare with log2 and readCount
  • RNA-seq bins: CPM
  • ATAC-seq: CPM for the shifted-alignment template

Safety notes: scripts and network use

This skill runs scripts and uses the network. Its scripts and commands read and write files on your machine, and the network is needed only for installation or for retrieving public input data, according to the skill's source.

Read the generated scripts before running them, and check them against your reference and file paths. The validation script does not establish assembly identity or biological suitability, so outputs from clinical or diagnostic work need expert review.

  • Review generated shell scripts before running them
  • Confirm the reference genome and effective genome size match your data
  • Do not use outputs for clinical or diagnostic decisions without expert review

What people use it for

  • Turn ChIP-seq BAM files into normalized bigWig tracks for a genome browser
  • Run a ChIP-seq QC pass with correlation, PCA and fingerprint plots across replicates
  • Draw a heatmap or profile of signal around TSS or called peaks
  • Build a reproducible bash script for a standard ATAC-seq or RNA-seq coverage workflow

Frequently asked questions

Is the deepTools skill free to use?

The listing states the licence as MIT. The SKILL.md header, however, says BSD license, so check the licence file in the source repository before you rely on either. deepTools itself is installed separately from PyPI or bioconda.

What do I need to use the deepTools skill?

You need Python 3.12 or newer (the skill recommends 3.13), deepTools with pysam and pyBigWig, and samtools for sorting and indexing. You also need alignment files (BAM) or other sequencing data. uv or conda is optional for setting up the environment.

Can I use the deepTools skill without command-line experience?

Not easily. The skill assumes you have alignment files and a Python environment. Users without sequencing data or command-line experience will find it hard to use productively.

How do I install the skill in Claude?

Download the ZIP and unzip it into ~/.claude/skills/ for personal use or into a project's .claude/skills/ folder. In the Claude apps, upload the ZIP under Settings → Capabilities → Skills.

How does deepTools compare with other tools for coverage tracks?

This guide does not compare deepTools with other tools, because the source material does not cover alternatives. Check the skill's reference files and your lab's pipeline needs before choosing.

Why does my deepTools analysis fail or give odd results?

Check that your reference genome matches the effective genome size you used, that your BAM files are sorted and indexed, and that the validation script passes. The skill also notes that RPGC without extension failed on shifted ATAC-seq BAM files in its tested release, so use CPM for that template.

New agent skills, repos and AI tools — every week

The best new skills and open-source projects, the fastest-growing repos, and one free prompt. Five minutes to read. Unsubscribe anytime.

Our verdict

This skill suits bioinformaticians and students who already work with BAM files and want an agent to write and check deepTools commands for ChIP-seq, RNA-seq and ATAC-seq visualization and QC. It is not suited to clinical or diagnostic decisions, to users without sequencing data or command-line experience, or to gene-level RNA-seq quantification, which needs an annotation-aware workflow.

Download the skill