African Natural Products Database Annotation Pipeline

Annotation of ANPDB compounds using the Ersilia Model Hub. This project is part of an ongoing collaboration between Prof. Ntie-Kang's Centre for Drug Discovery at the University of Buea (UB-CeDD) and the Ersilia Open Source Initiative.

The purpose of this pipeline is to annotate the ANPDB with a few selected models from the Ersilia Model Hub. The pipeline starts with chemical structures (SMILES strings) and it returns a table with multiple calculations properties and predictions.

At a high level, we calculate the following properties:

• Physicochemical properties
• Synthetic accessibility properties
• ADMET properties
• Bioactivity prediction against some pathogens such as Plasmodium falciparum and Mycobacterium tuberculosis.
• Chemical space exploration components for 2D visualization

There is an app associated with this pipeline. Click the link below to access it:

• ANPDB Annotation Demo App

Installation

We recommend that you create a Conda environment to run the pipeline. A few dependencies are also necessary, such as rdkit and the standardiser.

conda create -n anpdb-annotation python=3.10
conda activate anpdb-annotation
cd anpdb-annotation
pip install -r requirements.txt

Then you need to make sure that the Ersilia CLI is appropriately installed. Docker should be active in your system.

Fetch the following models:

ersilia fetch

How to run the the pipeline

To annotate ANPDB, simply run the following:

bash run_pipeline.sh

An anpdb_annotated.tsv file will be stored in the results/ folder.

How to interpret the results table