GraphDRP

GraphDRP model for drug response prediction (DRP).

Dependencies

Check conda_env_py37.sh

ML framework:

• Torch
• Pytorch_geometric -- for graph neural network (GNN)

IMPROVE lib:

• improve_lib
• candle_lib -- improve lib dependency

Dataset

Benchmark data for cross-study analysis (CSA) can be downloaded from this site. The required data tree is shown below:

csa_data/raw_data/
├── splits
│   ├── CCLE_all.txt
│   ├── CCLE_split_0_test.txt
│   ├── CCLE_split_0_train.txt
│   ├── CCLE_split_0_val.txt
│   ├── CCLE_split_1_test.txt
│   ├── CCLE_split_1_train.txt
│   ├── CCLE_split_1_val.txt
│   ├── ...
│   ├── GDSCv2_split_9_test.txt
│   ├── GDSCv2_split_9_train.txt
│   └── GDSCv2_split_9_val.txt
├── x_data
│   ├── cancer_copy_number.tsv
│   ├── cancer_discretized_copy_number.tsv
│   ├── cancer_DNA_methylation.tsv
│   ├── cancer_gene_expression.tsv
│   ├── cancer_miRNA_expression.tsv
│   ├── cancer_mutation_count.tsv
│   ├── cancer_mutation_long_format.tsv
│   ├── cancer_mutation.parquet
│   ├── cancer_RPPA.tsv
│   ├── drug_ecfp4_nbits512.tsv
│   ├── drug_info.tsv
│   ├── drug_mordred_descriptor.tsv
│   └── drug_SMILES.tsv
└── y_data
    └── response.tsv

Note! ./data contains data files that were used to train and evaluate the GraphDRP for the original paper.

Source codes

• graphdrp_preprocess_improve.py: creates data files for drug resposne prediction (DRP)
• graphdrp_train_improve.py: trains the GraphDRP model
• graphdrp_infer_improve.py: runs inference with the trained GraphDRP model
• graphdrp_params.txt: parameter file

Step-by-step running

1. Clone the repo

git clone https://github.com/JDACS4C-IMPROVE/GraphDRP/tree/develop
cd GraphDRP

2. Download benchmark data

sh ./download_csa.sh

This will download the cross-study benchmark data into ./csa_data/.

3. Set computational environment

• Install dependencies (check conda_env_py37.sh)
• Set the required environment variables to point towards the data folder and improve lib. You need to download the improve lib repo (follow this repo for more info https://github.com/JDACS4C-IMPROVE/IMPROVE).

export IMPROVE_DATA_DIR="./csa_data/"
export PYTHONPATH=$PYTHONPATH:/lambda_stor/data/apartin/projects/IMPROVE/pan-models/IMPROVE

4. Preprocess benchmark data (raw data) to construct model input data (ML data)

python graphdrp_preprocess_improve.py

Generates:

• three model input data files: train_data.pt, val_data.pt, test_data.pt
• three tabular data files, each containing y data (responses) and metadata: train_y_data.csv, val_y_data.csv, test_y_data.csv

ml_data
└── GDSCv1-CCLE
    └── split_0
        ├── processed
        │   ├── test_data.pt
        │   ├── train_data.pt
        │   └── val_data.pt
        ├── test_y_data.csv
        ├── train_y_data.csv
        ├── val_y_data.csv
        └── x_data_gene_expression_scaler.gz

5. Train GraphDRP model

python graphdrp_train_improve.py

Trains GraphDRP using the processed data: train_data.pt (training), val_data.pt (for early stopping).

Generates:

• trained model: model.pt
• predictions on val data (tabular data): val_y_data_predicted.csv
• prediction performance scores on val data: val_scores.json

out_models
└── GDSCv1
    └── split_0
        ├── best -> /lambda_stor/data/apartin/projects/IMPROVE/pan-models/GraphDRP/out_models/GDSCv1/split_0/epochs/002
        ├── epochs
        │   ├── 001
        │   │   ├── ckpt-info.json
        │   │   └── model.h5
        │   └── 002
        │       ├── ckpt-info.json
        │       └── model.h5
        ├── last -> /lambda_stor/data/apartin/projects/IMPROVE/pan-models/GraphDRP/out_models/GDSCv1/split_0/epochs/002
        ├── model.pt
        ├── out_models
        │   └── GDSCv1
        │       └── split_0
        │           └── ckpt.log
        ├── val_scores.json
        └── val_y_data_predicted.csv

6. Run the trained model in inference mode on test data

python graphdrp_infer_improve.py This script uses the processed data and the trained model to evaluate performance.

Generates:

• predictions on test data (tabular data): test_y_data_predicted.csv
• prediction performance scores on test data: test_scores.json

out_infer
└── GDSCv1-CCLE
    └── split_0
        ├── test_scores.json
        └── test_y_data_predicted.csv