LRMAHpan: Peptide-HLA presentation Modeling

This repository contains the codes for the paper [LRMAHpan: a novel tool for multi-allelic HLA presentation prediction using Resnet-based and LSTM-based neural network]

LRMAHpan (Peptide-HLA Presentation Modeling) is a deep learning-based model to predict the Peptide-HLA presentation, including two submodels LRMAHpan AP (Antigen Processing) and LRMAHpan BA (Binding Affinity).

AP Model only takes the primary sequences of peptides as input. BA Model integrates both peptide sequences and 6 HLA types as input.

Dependency

1. Install basic packages using:

   # [Optional] Create a new environment and activate it
   conda create -n LRMAHpan python=3.7.13
   conda activate LRMAHpan
   
   # Install Pytorch packages (for CUDA 11.3)
   conda install pytorch==1.10.0 torchvision==0.11.0 -c pytorch
   conda install cudatoolkit=11.3.1 -c conda-forge 
   # Install other packages
   pip install torchtext==0.11.0
   pip install spacy==3.4.1
   pip install pandas==1.3.5
   pip install scikit-learn==1.0.2
   

   Note: Change the Pytorch version to be compatible with your CUDA version.

Inference

Predict

1. Put your input Peptide-HLA sequence pairs in the uploaded/multiple_query.csv file. The peptide are represented by their sequences in the following format:

   peptide	A1	A2	B1	B2	C1	C2
   KLEDLERDL	HLA-A*01:01	HLA-A*02:01	HLA-B*01:01	HLA-B*37:01	HLA-C*05:02	HLA-A*07:01
   AVLEOSGFRK	HLA-A*01:01	HLA-A*02:01	HLA-B*01:01	HLA-B*37:01	HLA-C*05:02	HLA-A*07:01

2. If only one HLA type（such as HLA-A*02:01） is provided, the input data format is as follows:

   peptide	A1	A2	B1	B2	C1	C2
   KLEDLERDL	HLA-A*02:01	HLA-A*02:01	HLA-A*02:01	HLA-A*02:01	HLA-A*02:01	HLA-A*02:01
   AVLEOSGFRK	HLA-A*02:01	HLA-A*02:01	HLA-A*02:01	HLA-A*02:01	HLA-A*02:01	HLA-A*02:01

3. Run

   python predict_res.py -t uploaded/multiple_query.csv -o ./download/result_multipe.csv
   

4. The predicted peptide-HLA presentation scores are in the /download/result_multipe.csv. The PS column in the file represent the predicted presentation scores (probabilities) of the peptide-HLA pair.

Contact

If you have any questions, please contact us at [email protected]