Data and Code for IJCAI'21 paper - SafeDrug

Folder Specification

data/
- get_SMILES.py: which is a crawler, given the drug ATC-4 level code (four digit, e.g., 'A01A'), our crawler returns (a set of) SMILES strings of that ATC-4 class (crawled from drugbank). This file needs atc2rxnorm.pkl (which maps ATC-4 code to rxnorm code and then query to drugbank), generated from ndc2rxnorm_mapping.txt and ndc2atc_level4.csv.
- ddi_mask_H.py: This file will output the bipartite structure of drug molecules and the fragments/substructures.
- processing.py: our data preprocessing file.
  - Input (extracted from external resources)
    - PRESCRIPTIONS.csv: the prescription file from MIMIC-III raw dataset
    - DIAGNOSES_ICD.csv: the diagnosis file from MIMIC-III raw dataset
    - PROCEDURES_ICD.csv: the procedure file from MIMIC-III raw dataset
    - idx2SMILES.pkl: drug ID (we use ATC-4 level code to represent drug ID) to drug SMILES string dict (This file is created by get_SMILES.py, due to the change of drug bank web structure, it may need updates)
    - ndc2atc_level4.csv: this is a NDC-RXCUI-ATC5 file, which gives the mapping information
    - drug-atc.csv: this is a CID-ATC file, which gives the mapping from CID code to detailed ATC code (we should truncate later)
    - ndc2rxnorm_mapping.txt: rxnorm to RXCUI file
    - drug-DDI.csv: this a large file, could be downloaded from https://drive.google.com/file/d/1mnPc0O0ztz0fkv3HF-dpmBb8PLWsEoDz/view?usp=sharing
  - Output
    - data_final.pkl: intermediate result
    - ddi_A_final.pkl: ddi matrix
    - ddi_matrix_H.pkl: H mask structure (This file is created by ddi_mask_H.py)
    - ehr_adj_final.pkl: used in GAMENet baseline (if two drugs appear in one set, then they are connected)
    - (important) records_final.pkl: 100 patient visit-level record samples. Under MIMIC Dataset policy, we are not allowed to distribute the datasets. Practioners could go to https://physionet.org/content/mimiciii/1.4/ and requrest the access to MIMIC-III dataset and then run our processing script to get the complete preprocessed dataset file.
    - voc_final.pkl: diag/prod/med index to code dictionary
- follow this figure and figure out the data preprocessing flow
src/
- SafeDrug.py: our model
- Epoch_49_TARGET_0.06_JA_0.5183_DDI_0.05854.model: the model we trained on the training set
- baselines:
  - GAMENet.py
  - DMNC.py: there are some bg issues for the latest dnc package, please refer to the original DMNC repo https://github.com/thaihungle/DMNC
  - Leap.py
  - Retain.py
  - ECC.py
  - LR.py
- setting file
  - model.py
  - util.py
  - layer.py

Step 1: Data Processing

Go to https://physionet.org/content/mimiciii/1.4/ to download the MIMIC-III dataset (You may need to get the certificate)
```
cd ./data
wget -r -N -c -np --user [account] --ask-password https://physionet.org/files/mimiciii/1.4/
```

go into the folder and unzip three main files

cd ./physionet.org/files/mimiciii/1.4
gzip -d PROCEDURES_ICD.csv.gz # procedure information
gzip -d PRESCRIPTIONS.csv.gz  # prescription information
gzip -d DIAGNOSES_ICD.csv.gz  # diagnosis information

download the DDI file and move it to the data folder download https://drive.google.com/file/d/1mnPc0O0ztz0fkv3HF-dpmBb8PLWsEoDz/view?usp=sharing
```
mv drug-DDI.csv ./data
```

processing the data to get a complete records_final.pkl

cd ./data
vim processing.py

# line 294~296
# med_file = './physionet.org/files/mimiciii/1.4/PRESCRIPTIONS.csv'
# diag_file = './physionet.org/files/mimiciii/1.4/DIAGNOSES_ICD.csv'
# procedure_file = './physionet.org/files/mimiciii/1.4/PROCEDURES_ICD.csv'

python processing.py

run ddi_mask_H.py to get the ddi_mask_H.pkl
```
python ddi_mask_H.py
```

Step 2: Package Dependency

first, install the rdkit conda environment

conda create -c conda-forge -n SafeDrug  rdkit
conda activate SafeDrug

then, in SafeDrug environment, install the following package

pip install scikit-learn, dill, dnc

Note that torch setup may vary according to GPU hardware. Generally, run the following

pip install torch

If you are using RTX 3090, then plase use the following, which is the right way to make torch work.

python3 -m pip install --user torch==1.8.0+cu111 torchvision==0.9.0+cu111 torchaudio==0.8.0 -f https://download.pytorch.org/whl/torch_stable.html

Finally, install other packages if necessary

pip install [xxx] # any required package if necessary, maybe do not specify the version, the packages should be compatible with rdkit

Here is a list of reference versions for all package

pandas: 1.3.0
dill: 0.3.4
torch: 1.8.0+cu111
rdkit: 2021.03.4
scikit-learn: 0.24.2
numpy: 1.21.1

Let us know any of the package dependency issue. Please pay special attention to pandas, some report that a high version of pandas would raise error for dill loading.

Step 3: run the code

python SafeDrug.py

here is the argument:

usage: SafeDrug.py [-h] [--Test] [--model_name MODEL_NAME]
               [--resume_path RESUME_PATH] [--lr LR]
               [--target_ddi TARGET_DDI] [--kp KP] [--dim DIM]

optional arguments:
  -h, --help            show this help message and exit
  --Test                test mode
  --model_name MODEL_NAME
                        model name
  --resume_path RESUME_PATH
                        resume path
  --lr LR               learning rate
  --target_ddi TARGET_DDI
                        target ddi
  --kp KP               coefficient of P signal
  --dim DIM             dimension

If you cannot run the code on GPU for SafeDrug.py, just change line 101, "cuda" to "cpu" and change line 126 to

model.load_state_dict(torch.load(open(args.resume_path, 'rb'), map_location=torch.device('cpu')))

Citation

@inproceedings{yang2021safedrug,
    title = {SafeDrug: Dual Molecular Graph Encoders for Safe Drug Recommendations},
    author = {Yang, Chaoqi and Xiao, Cao and Ma, Fenglong and Glass, Lucas and Sun, Jimeng},
    booktitle = {Proceedings of the Thirtieth International Joint Conference on
               Artificial Intelligence, {IJCAI} 2021},
    year = {2021}
}