How rinaldi-rpy/smallruminantcadd works
Automated CADD Pipeline v2 is a configurable Snakemake workflow for building and validating variant-deleteriousness models and producing CADD scores for non-human species. Its broad workflow runs from reference/alignment preparation and variant-set generation through annotation, feature preparation, model training, and score generation; it also supports regional evaluation and conservation annotation generation. The README documents these capabilities, while sampled source confirms selected processing and modeling behavior. The exact rule-level wiring is not fully visible in the excerpts, so the graph keeps orchestration links broad rather than asserting unverified helper-to-helper calls.
Main entry points: Snakefile Snakemake workflow · annotate_variants.smk Variant annotation · cadd_score.smk CADD score generation · resources Reference and MSA
Components 19
Genome and Ancestry
- Reference and MSA
resources - Workflow configuration
config - Snakemake workflow
workflow/Snakefile - Ancestral extraction
workflow/rules/alignment_processing.smk - Coverage statistics
workflow/scripts/alignment_processing/report_ancestral_coverage.py
Variant Sets
- Derived variants
workflow/rules/derive_variants.smk - Simulated variants
workflow/rules/simulate_variants.smk - Population variants
Annotations
- Conservation data
workflow/rules/generate_annotations.smk - Variant annotation
workflow/rules/annotate_variants.smk - Annotation processors
workflow/scripts/annotate_variants
Modeling and Scores
- CADD score generation
workflow/rules/cadd_score.smk - Score processing
workflow/scripts/cadd_score - Feature preparation
workflow/scripts/annotate_variants/data_preparation.py - Train and test
workflow/rules/train_test_model.smk - Model fitting
workflow/scripts/train_test_model/train_model.py
Evaluation
- Regional evaluation
workflow/scripts/train_test_model/average_auc_by_region.py - Pipeline outputs
results
Other components
- Pipeline user
Connections 30
- Pipeline user → Snakemake workflow: runs
- Snakemake workflow → Workflow configuration: reads settings
- Snakemake workflow → Ancestral extraction: dispatches
- Reference and MSA → Ancestral extraction: provides sequences
- Ancestral extraction → Coverage statistics: coverage input
- Snakemake workflow → Conservation data: dispatches
- Reference and MSA → Conservation data: provides alignment
- Snakemake workflow → Derived variants: dispatches
- Reference and MSA → Derived variants: provides reference
- Population variants → Derived variants: provides variants
- Snakemake workflow → Simulated variants: dispatches
- Reference and MSA → Simulated variants: provides reference
- Snakemake workflow → Variant annotation: dispatches
- Derived variants → Variant annotation: variant input
- Simulated variants → Variant annotation: variant input
- Variant annotation → Annotation processors: invokes
- Conservation data → Variant annotation: annotation input
- Snakemake workflow → Train and test: dispatches
- Variant annotation → Feature preparation: annotated input
- Feature preparation → Train and test: prepared features
- Train and test → Model fitting: invokes
- Snakemake workflow → CADD score generation: dispatches
- Model fitting → CADD score generation: trained model
- Variant annotation → CADD score generation: annotated variants
- CADD score generation → Score processing: invokes
- Snakemake workflow → Regional evaluation: dispatches
- Train and test → Regional evaluation: test predictions
- Regional evaluation → Pipeline outputs: writes reports
- CADD score generation → Pipeline outputs: writes scores
- Coverage statistics → Pipeline outputs: writes statistics