Skip to content

ortholog_generation.md

Latest commit

 

History

History
59 lines (43 loc) · 2.25 KB

ortholog_generation.md

File metadata and controls

59 lines (43 loc) · 2.25 KB
Raw

Ortholog Generation

Across-species analyses in amalgkit (cstmm, csca) require an orthogroups.tsv (or Orthogroups.tsv) table that maps gene IDs across different species. MetaInformAnt provides an automated workflow to generate this table using OrthoDB v12 data.

Workflow Overview

Generating ortholog mappings manually is error-prone and time-consuming. MetaInformAnt automates this through a single orchestrator that syncs the required 5GB+ of genomic databases and extracts the relevant taxonomy IDs.

graph TD
    A[amalgkit_*.yaml] --> B(generate_orthologs.py)
    B --> C{Databases Exist?}
    C -- No --> D[Stream from data.orthodb.org]
    C -- Yes --> E[Extract Taxon IDs]
    E --> F(create_ortholog_table.py)
    F --> G[orthogroups.tsv]
Loading

Core Components

1. generate_orthologs.py

The top-level orchestrator.

  • Discovery: Automatically parses all config/amalgkit/amalgkit_*.yaml files to find configured taxon_ids.
  • Synchronization: Downloads required OrthoDB v12 files if missing from .cache/orthodb/.
  • Execution: Triggers the extraction logic.
# Automated generation for all configured species
python3 scripts/rna/generate_orthologs.py

2. create_ortholog_table.py

The extraction engine.

  • Filters the massive OrthoDB gene-to-OG tables for the specific taxonomy IDs of interest.
  • Produces the standard orthogroups.tsv format expected by amalgkit cstmm.

Configuration

Species are included in ortholog generation if they have a taxon_id defined in their amalgkit YAML:

# config/amalgkit/amalgkit_acromyrmex_echinatior.yaml
taxon_id: 103372
...

Data Sources

The workflow uses OrthoDB v12.2 as the source of truth, specifically:

  • odb12v2_genes.tab.gz: Gene ID mappings.
  • odb12v2_OG2genes.tab.gz: Orthogroup to Gene ID mappings.

These are retrieved securely from https://data.orthodb.org/v12/.

Validation

The ortholog generation workflow is validated via:

  • Automated Tests: test_ortholog_generation.py verifies correct Taxon ID extraction from configuration files.
  • Output Integrity: Verified by the cstmm step, which fails if the ortholog table contains invalid IDs or mismatched headers.