Guanghui Li, Peihao Bai, Jiao Chen, Cheng Liang, Identifying virulence factors using graph transformer autoencoder with ESMFold-predicted structures, Computers in Biology and Medicine, 2024, 170: 108062.

Code

Environment Requirement

The code has been tested running under Python 3.8.16. The required packages are as follows:

• numpy == 1.23.5
• numpy-base == 1.23.5
• openfold == 1.0.0
• networkx == 3.1
• scipy == 1.10.1
• pytorch == 1.13.1
• pytorch-lightning == 1.5.10
• pytorch-cuda ==11.6

Files

1.dataset: dataset.fasta store the protein sequence information of the training set, validation set, and test set

2. src: a.Model.py：the GTAE-VF framework； b.graphT.py: the graph transformer framework; c.main.py: training model saves the optimal parameters of the model.