Hi everyone,

I'd like to share a lightweight CLI tool I developed to help with large-scale TE (Transposable Element) annotation projects: EDTA-Audit-Tool.

GitHub: https://github.com/dogdogdoghead/EDTA-Audit-Tool

The Problem

When running EDTA on hundreds of genomes (e.g., pangenome projects), checking the logs one by one is painful. A simple ls -lh check isn't enough because:

1. Failed jobs might still produce empty folders.
2. "Finished" jobs might produce suspiciously small library files (e.g., a 2KB LTR library in a plant genome is likely a failure).
3. Different species (Plants vs. Fungi) have vastly different TE compositions.

The Solution

EDTA-Audit-Tool is a Bash script that performs biology-aware Quality Control. It scans your output directories and flags samples that fail to meet species-specific thresholds.

Key Features

• Species-Aware Thresholds:
  • Plant Mode: Expects high TE content (>20%) and large LTR libraries (>500KB).
  • Fungi Mode: Permissive. Allows missing SINEs/Helitrons and low TE content.
  • Animal Mode: Adjusted for LINE/SINE dominance.
• Hybrid Search: Works even if your output directory structure is messy (nested or flat).
• Pipeline Ready: Generates a clean failed_list.txt that you can pipe directly into a re-run loop.

Usage Example

# Check plant genomes (Strict QC)
./edta_audit.sh -d ./rice_genomes -t plant

Output:

[1] Oryza_sativa 42.5% | LTR:4.8M TIR:3.7M LINE:39K Helitron:322K SINE:26K

[2] Failed_Sample_X 0.00% | LTR:Miss TIR:Miss LINE:Miss Helitron:Miss SINE:Miss