lxlsu / fragrow Goto Github PK

• dataset without protein data：0507ligand_pro_15cutways.txt
  The first column is an Smiles sequence of the intact molecule, and the second column is an Smiles sequence of the cropped molecule
• dataset with protein data：0525ligand_pro_15cutways.txt
  The first column is an Smiles sequence of the intact molecule, the second column is an Smiles sequence of the clipped molecule, and the third column is a protein sequence.

// run the code, each model code is in the corresponding folder
run LSTM_molecule0507.py

Take the * ./LSTM folder as an example:

• ./r1: Experimental results
• ./r1/parameters.txt : input hyperparameters and loss output
• ./r1/evaluate.txt: some indicators of prediction results
• ./r1/encoder.pth/decoder.pth: contains the trained model
• ./LSTMresults : Model prediction results