LRMAHpan

LRMAHPan: a novel tool for multi-allelic HLA presentation prediction using Resnet-based and LSTM-based neural networks

INSTALL

install pytorch gpu

install torchtext

install spacy

install pandas

install numpy

DATA

Big data for training and 10 independent test datasets are stored in the data.zip under github's master path: https://github.com/Luckysoutheast/LRMAHpan/tree/master. You can learn more about the details of the original data in this link.

USE

python predict_res.py –t testfile.csv –o outfile.csv

Test set should be a comma-separated CSV files. The files should have the following columns (with headers): peptide, label, A1, A2, B1, B2, C1, C2(the label column is not required when you predict)

For example:

python predict_res.py -t ./uploaded/multiple_query.csv -o result_multiple_query.csv

Result

The results of the stand-alone test set are stored in the test_result.zip file