Data Centric Molecular Analysis and Evaluation of Hepatocellular Carcinoma Therapeutics Using Machine
Rengul Cetin-Atalay1, Deniz Cansen Kahraman2, Esra Nalbat3
1Section of Pulmonary and Critical Care Medicine, University of Chicago, Chicago, IL, 60637, USA. rengul@uchicago.edu.
Journal of Gastrointestinal Cancer
|December 15, 2021
Summary
This study uses artificial intelligence (AI) to accelerate drug discovery for hepatocellular carcinoma (HCC). AI models identified potential new drug candidates by analyzing biological data and molecular similarities to existing HCC drugs.
Area of Science:
- Biomedicine
- Computational Biology
- Drug Discovery
Background:
- Computational approaches reduce drug development time and cost.
- Artificial intelligence (AI) techniques are increasingly applied in biomedicine.
Purpose of the Study:
- To conduct a data-driven evaluation of potential hepatocellular carcinoma (HCC) therapeutics using AI-assisted drug discovery and repurposing.
- To explore AI's role in identifying novel therapeutic strategies for complex diseases.
Main Methods:
- Integrated HCC-related biological entities (genes, proteins, pathways, drugs) into a knowledge graph using the CROssBAR system.
- Employed deep learning models (DEEPScreen, MDeePred) for drug-target interaction prediction.
- Utilized 2-D embedding of protein ligands to identify and repurpose small molecule inhibitors based on molecular similarity.
Main Results:
- System-level evaluation identified critical genes and pathways for HCC targeting.
- Predicted novel bioactive drugs and compounds for selected HCC targets.
- Identified potential drug candidates by repurposing small molecule inhibitors based on molecular similarity to known HCC drugs.
Conclusions:
- The data-driven analysis provides a roadmap for proposing early-stage potential inhibitors for HCC and other complex diseases.
- These findings can guide subsequent in silico and experimental validation of therapeutic candidates.
