Adhesion Protein Predictor

A bioinformatics tool for predicting adhesion proteins from FASTA formatted sequence files.

Requirements

This will run with pytorch on CPUs but will be much faster if on a GPU system with torchvision installed having CUDA bindings.

Usage

Training

I have built a set of FLO, ALS1 related proteins from Saccharomyces and Candida for starters. This seems to have some reasonable power.

python scripts/training.py --positive training/positive --negative training/negative

Application

You can run on a single file at a time and produce a report for each query file

mkdir -p query
pushd query
curl -O https://fungidb.org/a/service/raw-files/release-68/CalbicansSC5314/fasta/data/FungiDB-68_CalbicansSC5314_AnnotatedProteins.fasta
curl -O https://fungidb.org/a/service/raw-files/release-68/CneoformansJEC21/fasta/data/FungiDB-68_CneoformansJEC21_AnnotatedProteins.fasta
curl -O https://fungidb.org/a/service/raw-files/release-68/Spombe972h/fasta/data/FungiDB-68_Spombe972h_AnnotatedProteins.fasta
popd
for qorg in $(ls query/*.fasta)
do
   python scripts/predict.py --input $qorg --output $(basename $qorg .fasta).adhesion_predict.csv
done

You can run on a single folder and all results will be combined in a single file. It will look for all .fasta, .fa, .pep, .aa with or without .gz extensions.

python scripts/predict.py --input query --output Combinedquery_adhesion_predict.csv

Development Setup

1. Install Dependencies:

   pip install -r requirements.txt
   

2. Setup Pre-commit Hooks:

   pre-commit install
   

Author

Jason Stajich, jason.stajichucr.edu

Code was developed with support from opencode.ai, co-pilot and various code models for setting up classifier framework