dbCAN-Sum

Script to aggregate multiple result files from the dbCAN3 web server's automated Carbohydrate-active enzyme ANnotation for my project.

Overview

This tool processes and aggregates dbCAN3 annotation files for carbohydrate-active enzymes (CAZymes) across multiple species or samples.

Input Files

• Source: Test input files are generated from the dbCAN3 web server (link here).
• Preparation: Each input file is the overview.txt output file, renamed to correspond to the species or sample name (e.g., species1.txt, species2.txt).
• Configuration: dbCAN3 configured for testing with:
  • Input: The test species' total protein file in .faa format.
  • Databases: Selected HMMER_dbCAN, HMMER_dbCAN_sub, and DIAMOND databases for the prediction.

Aggregation Rules

The summarization process applies the following rules:

Tool Filtering

• Only genes with results from at least two tools (from HMMER_dbCAN, HMMER_dbCAN_sub, DIAMOND) are considered.

Annotation Extraction

• If a gene has multiple annotations, the annotation with the highest occurrence across the three tools is recorded.
  • Example: GT54(82-362)+CBM94(374-526), GT54_e0, and CBM94+GT54 would be recorded as GT54 since it appears most frequently.
• In cases where multiple annotations have the same occurrence, the first annotation appearing in HMMER_dbCAN_sub is used as the record.
• Note: This rule may differ from certain workflows, which may recommend recording all CAZyme predictions as separate entries.

Summarization

• Counts for major categories and families are aggregated per species or sample, using the renamed .txt files as identifiers.

Usage

To run the script, use the following command:

python dbCAN_statistics_multi_en.py -in '*.txt'