A learning-based tool to predict antimicrobial resistance and identify multimodal AMR-related genomic features from bacterial strain genomes
git clone https://github.com/liaoherui/StrainAMR.git
cd StrainAMR
unzip Test_genomes.zip
unzip localDB.zip
unzip Benchmark_features.zip
conda env create -f strainamr.yaml
conda activate strainamr
sh download_ps.sh
python install_rebuild_ps.py
(Important) Add environment variable for running phenotypeseeker:
Open the bashrc file:
vi ~/.bashrc
Add this line:
export PATH=$PATH:<yout installation directory>/PhenotypeSeeker/.PSenv/bin
For example, if my installation dir is /home/ray/StrainAMR, then it should be
export PATH=$PATH:/home/ray/StrainAMR/PhenotypeSeeker/.PSenv/bin
Finally:
source ~/.bashrc
Test your installation：
python StrainAMR_build_train.py -h
python StrainAMR_build_test.py -h
python StrainAMR_model_train.py -h
python StrainAMR_model_pred.py -h
Test the tool with the demo data:
sh test_run.sh
Rerun the 3-fold cross validation experiment:
sh batch_train_3fold_exp.sh