What is this pDILI_v1?

pDILI_v1 is a python package that allows users to predict the association of Drug-Induced Liver Injury (DILI) of a small molecule (1 = RISKy, 0 = Non-RISKy) and also visualize the molecule.

pDILI_v1 stands for predictor of Drug-Induced Liver Injury.

It is a -based Web Application.

It is also an online tool hosted on Google Colab .

This tool also provides Graphical User Interface (GUI) which can be run through Anaconda environment (in Windows).

Note: This notebook is a supplementary material of the paper "pDILI_v1: A Machine Learning-Based Tool for Predicting Drug-Induced Liver Injury (DILI) Integrating Chemical Space Analysis and Molecular Fingerprints" (manuscript under preparation).

For Web Application Users:

The pDILI_v1 web application can be used by following This Link.

Just enter a SMILES string to predict its DILI-Risk (RISKy / Non.RISKy)!

---

Example SMILES:

(a) Sorbitol: C(C(C(C(C(CO)O)O)O)O)O

(b) Almotriptan: CN(C)CCC1=CNC2=C1C=C(C=C2)CS(=O)(=O)N3CCCC3

(c) Imatinib: Cc1ccc(NC(=O)c2ccc(CN3CCN(C)CC3)cc2)cc1Nc1nccc(-c2cccnc2)n1

---

For Google Colab Users:

Please follow these three steps before running this notebook.

1: Download the two csv files provided herewith (named '1_train_pDILI.csv' and '2_test_pDILI.csv') and create a folder named "pDILI_v1". Move these two csv files in to the folder pDILI_v1.

or Download the folder named "pDILI_v1" Directly Download.

2: Upload this folder (pDILI_v1) in your Google Drive. Copy this path. Make sure that ''1_train_pDILI.csv' and '2_test_pDILI.csv' are present in that folder pDILI_v1.

3: and execute it to predict the DILI RISK property of the query molecule as well as visualize the Applicability domain (AD).

---

For Graphical User Interface (GUI) Users:

1: Installation of Anaconda Downalod Anaconda & Install it.

If you already have Anaconda in your system then skip the Installation.

2: Open Anaconda Prompt

3: Then activate environment (for example: pDILI_v1). You can copy and run this command directly in your terminal:

conda create -n pDILI_v1 python=3.9  #for the first time only

You can copy and run this command directly in your terminal:

conda activate pDILI_v1

3(for the first time only). Then install the required packages (one time only) You can copy and run this command directly in your terminal:

conda install -c anaconda tk #for the first time only

conda install -c conda-forge rdkit pandas scikit-learn pillow #for the first time only

#or

pip install rdkit pandas scikit-learn pillow #for the first time only

pip install matplotlib numpy mordred #for the first time only

4. Go to your working directory

cd yourpath #for example D:\DILI_Amin\pDILI_v1

Ensure that the training, test set, the pDILI_logo and the pDILI_v1.py should be present in the your working directory (pDILI_v1).

Then run the command

python pDILI_v1.py

---

Bugs: If you encounter any bugs, please report the issue to Dr. Sk. Abdul Amin.