Bioinformatics Skills Gateway: 400+ Genomics & Computational Biology Skills

Overview

This reference describes the gateway to two open-source bioinformatics skill libraries that are fetched on demand rather than bundled. The system indexes and retrieves domain-specific skills from external repositories when you need them, keeping the core lightweight. The gateway provides access to 418 total skills across two repositories, covering the full spectrum of computational biology.

When to Use

Use this capability when asked about bioinformatics, genomics, sequencing, variant calling, gene expression, single-cell analysis, protein structure, pharmacogenomics, metagenomics, phylogenetics, or any computational biology task. The gateway is designed for situations where you need validated tool parameters, pipeline scripts, or decision trees for bioinformatics workflows.

Capabilities

The gateway provides access to two repositories:

• bioSkills: 385 reference skills covering code patterns, parameter guides, and decision trees in Python, R, and CLI. These are cookbook-style references that show correct tool usage and parameter combinations. Each skill is stored as a SKILL.md file within a category directory.
• ClawBio: 33 runnable pipeline skills with executable scripts and reproducibility bundles. These include Python scripts and README files that can be executed directly. Many skills include a --demo flag for testing.

The skills cover the following domains: sequence fundamentals, read QC and alignment, variant calling and annotation, differential expression (bulk RNA-seq), single-cell RNA-seq, spatial transcriptomics, epigenomics, pharmacogenomics and clinical, population genetics and GWAS, metagenomics and microbiome, genome assembly and annotation, structural biology and chemoinformatics, proteomics, pathway analysis and gene networks, immunoinformatics, CRISPR and genome engineering, workflow management, specialized domains (alternative splicing, ecological genomics, epidemiological genomics, liquid biopsy, epitranscriptomics, metabolomics, flow cytometry, systems biology, RNA structure), data visualization and reporting, database access, experimental design, and machine learning for omics.

Prerequisites

Before using the skills, ensure the following are installed. The bioinformatics tools required for the pipelines are extensive, with approximately 8000 packages available across ecosystems and over 12000 tools in the broader bioinformatics landscape. Genomic data files can be very large, so ensure adequate disk space for reference genomes, SRA datasets, and indices.

Python Packages

# Python
pip install biopython pysam cyvcf2 pybedtools pyBigWig scikit-allel anndata scanpy mygene

R/Bioconductor Packages

# R/Bioconductor
Rscript -e 'BiocManager::install(c("DESeq2","edgeR","Seurat","clusterProfiler","methylKit"))'

CLI Tools (Ubuntu/Debian)

# CLI tools (Ubuntu/Debian)
sudo apt install samtools bcftools ncbi-blast+ minimap2 bedtools

CLI Tools (macOS)

# CLI tools (macOS)
brew install samtools bcftools blast minimap2 bedtools

Via Conda (Recommended)

# Or via Conda (recommended for reproducibility)
conda install -c bioconda samtools bcftools blast minimap2 bedtools fastp kraken2

Git

Git is required for cloning the repositories. Ensure Git is installed and accessible from the command line.

Fetching and Using Skills

Procedure: Fetch and Use a Bioinformatics Skill

Follow these steps in order:

1. Identify the domain and skill name from the skill index provided in the document. The index lists all available skills by category. For bioSkills, the index is organized by domain (e.g., variant-calling, single-cell). For ClawBio, the index lists pipeline names.
2. Clone the relevant repository using a shallow clone. Use the following commands exactly:

# bioSkills (reference material)
git clone --depth 1 https://github.com/GPTomics/bioSkills.git /tmp/bioSkills

# ClawBio (runnable pipelines)
git clone --depth 1 https://github.com/ClawBio/ClawBio.git /tmp/ClawBio

3. Read the specific skill file:

# bioSkills — each skill is at: <category>/<skill-name>/SKILL.md
cat /tmp/bioSkills/variant-calling/gatk-variant-calling/SKILL.md

# ClawBio — each skill is at: skills/<skill-name>/
cat /tmp/ClawBio/skills/pharmgx-reporter/README.md

4. Follow the fetched skill as expert reference material. These are not Hermes-format skills; treat them as domain guides with correct parameters, tool flags, and validated pipelines. For ClawBio skills, install dependencies before running:

pip install -r requirements.txt

Run this command from the cloned repository root.

Parameters

Parameter	Meaning	Required
skill domain	The broad area of bioinformatics (e.g., variant-calling, single-cell) from the index.	Yes
skill name	The specific skill within a domain (e.g., gatk-variant-calling, scrna-orchestrator).	Yes
repository URL	The GitHub URL for bioSkills or ClawBio, used for cloning.	Yes
clone depth	Use --depth 1 for shallow clone to save time.	No (default is full clone)

Skill Structure and Usage

bioSkills (Reference Material)

bioSkills are reference guides showing correct parameters and code patterns but are not executable pipelines. Each skill is stored as a SKILL.md file within a category directory. For example, the variant calling skill is at /tmp/bioSkills/variant-calling/gatk-variant-calling/SKILL.md. These files contain code pattern cookbooks with validated parameter combinations and decision trees. The structure follows a consistent format: domain, tool, parameters, and example usage. Do not attempt to run these files as scripts; they are reference material only.

ClawBio (Runnable Pipelines)

ClawBio skills are executable pipelines. Each skill has a README.md file and associated Python scripts. The README provides usage instructions, input requirements, and output descriptions. Many skills include a --demo flag that can be used to run them directly with test data. Before running any ClawBio pipeline, install the dependencies:

pip install -r requirements.txt

Run this command from the cloned repository root. The Python scripts are designed to be executed from the skills directory. For example, to run the pharmacogenomics reporter:

cd /tmp/ClawBio/skills/pharmgx-reporter
python pharmgx_reporter.py --demo

Constraints and Caveats

• Fetched skills are NOT in Hermes SKILL.md format; they use their own structure. bioSkills are code pattern cookbooks; ClawBio uses README files plus Python scripts.
• bioSkills are reference guides only. They show correct parameters and code patterns but are not executable pipelines. Do not attempt to run them as scripts.
• ClawBio skills are executable. Many have a --demo flag and can be run directly.
• Both repositories assume bioinformatics tools are installed. Check the prerequisites section before running any pipeline.
• For ClawBio, run pip install -r requirements.txt in the cloned repo first.
• Genomic data files can be very large. Be mindful of disk space when downloading reference genomes, SRA datasets, or building indices. The bioinformatics ecosystem contains approximately 8000 packages and over 12000 tools, so you may need significant storage for indices and reference data.
• Cloning without --depth 1 may take excessive time and disk space, especially for the full history of these repositories.
• The skill index is provided separately and must be consulted to identify the correct domain and skill name before cloning.

Failure Modes

Failure Mode	Description
Missing prerequisites	Tools not installed, leading to pipeline failures. Check the prerequisites section before running any skill.
Insufficient disk space	Large genomic data files can fill storage quickly. Ensure at least 10 GB free for reference data.
Cloning without --depth 1	Full clones may take excessive time and disk space. Always use --depth 1 unless you need the full history.
Attempting to run bioSkills as executable pipelines	They are reference only. Trying to execute SKILL.md files will fail.
Not installing ClawBio dependencies	Running skills without pip install -r requirements.txt will fail due to missing Python packages.
Incorrect skill path	Using the wrong path for the skill file will result in file not found errors. Verify the domain and skill name from the index.
Running ClawBio scripts from wrong directory	The Python scripts may have relative path dependencies. Always run them from the skill's directory.

Examples

• bioSkills: variant-calling/gatk-variant-calling/SKILL.md is a reference skill for GATK variant calling, showing correct parameters and code patterns. It includes best practices for read mapping, duplicate marking, base recalibration, and variant calling with GATK HaplotypeCaller.
• ClawBio: skills/pharmgx-reporter/README.md is a runnable pipeline for pharmacogenomics reporting from consumer genetic data. It includes executable scripts and can be tested with the --demo flag. The pipeline processes raw genotype data and generates a report of pharmacogenomic variants.
• ClawBio: skills/scrna-orchestrator/README.md is a runnable pipeline for single-cell RNA-seq analysis. It includes scripts for preprocessing, clustering, and visualization. Run with --demo to test on sample data.

Skill Index

The skill index is provided as a separate document that lists all available skills by domain. Consult this index to identify the correct domain and skill name before fetching. The index includes:

• For bioSkills: domain, skill name, and brief description of what the skill covers.
• For ClawBio: pipeline name, description, and whether it supports the --demo flag.

The index is organized alphabetically by domain and then by skill name. Use it to find the specific skill you need for your bioinformatics task.