Clear Cache

When a pipeline script or rule file changes, Snakemake may not detect the change automatically. This skill identifies which output directories depend on the changed file and removes them to force regeneration.

Usage

Invoke with an optional argument specifying the changed file path. If no argument is given, check git diff --name-only for recently modified workflow scripts.

Dependency Map

Use these known dependency chains to determine what to clear:

Header/Annotation scripts

If generate_header_lines.py, combine_beds_with_id.py, or annotation-related scripts changed:

code

rm -rf results/generate_annotation_headers/
rm -rf results/annotate_vcf_genomic_contexts/
rm -rf results/annotate_vcf_regions/
rm -rf results/combine_genomic_context_beds/
rm -rf results/variant_tables/

Variant table scripts

If generate_variant_parquet.py or variant classification scripts changed:

code

rm -rf results/variant_tables/
rm -rf results/var_counts/

Variant count scripts

If count_variants_by_genomic_context.py changed:

code

rm -rf results/var_counts/

Metrics scripts

If compute_bed_metrics.py or metrics scripts changed:

code

rm -rf results/genomic_context_metrics/

Exclusion scripts

If exclusion-related scripts changed:

code

rm -rf results/exclusions/

R cache

If R/schemas.R, R/cache.R, or R/data_loading.R changed:

code

rm -rf analysis/cache/

Rules

•Always show the user which directories will be deleted and get confirmation before proceeding
•After clearing, suggest running micromamba run -n q100-smk snakemake -n -- all to verify the DAG
•If the changed file doesn't match any known dependency, warn the user and suggest checking the Snakemake rule that uses that script