bioinformatics

Browse the skill index organized by domain, such as variant calling, RNA-seq, or metagenomics, then shallow-clone the bioSkills or ClawBio repository and read the specific skill file. bioSkills provides reference guides while ClawBio provides executable pipelines.

bioSkills contains 385 reference skills with code patterns, parameter guides, and decision trees in Python, R, and CLI. ClawBio contains 33 runnable pipeline skills with executable Python scripts, demo flags, and reproducibility bundles exporting report.md, commands.sh, and environment.yml.

Yes, many ClawBio skills are executable and include --demo flags for direct runs. First run pip install -r requirements.txt in the cloned repo, and ensure underlying bioinformatics tools like samtools or BWA are installed.

Common dependencies include Python packages like biopython, pysam, and scanpy, R/Bioconductor packages like DESeq2 and Seurat, and CLI tools like samtools, bcftools, and bedtools. Conda with the bioconda channel is recommended for reproducibility.

No, the fetched skills use their own structures and are not Hermes-format skills. Treat them as expert domain reference material containing validated parameters, correct tool flags, and tested pipelines rather than directly loadable skills.

The gateway only indexes external repositories, so it requires network access to clone them and a workstation with bioinformatics tools preinstalled. Genomic data files and reference genomes can be very large, so disk space must be managed carefully.