Recent Advances and Perspectives of Artificial Intelligence in the Discovery of Enzyme Inhibitors for Metabolic Syndrome: A Focus on Computational Design and Data Platforms

Read full text Summarise with VIRAT
Authors: Uriel Zagada Dominguez
Cited by: 0
References: 4

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

The discovery of enzyme inhibitors for Metabolic Syndrome (encompassing obesity and type 2 diabetes mellitus) is undergoing a paradigm shift driven by data science and computational power. This review article systematically analyzes recent literature (2019-2024), focusing on Artificial Intelligence (AI)-assisted rational drug design, the exploitation of ultra-large chemical spaces, and the underlying database infrastructures. We examine how Deep Learning architectures-specifically molecular Graph Neural Networks (GNNs), Transformers, and Generative Adversarial Networks (GANs)-have optimized the screening and de novo design of inhibitors targeting DPP-4, PTP1B, and Pancreatic Lipase. Our analysis indicates that integrating makeon-demand databases with AI platforms allows researchers to explore chemical spaces of up to 10 10 molecules, vastly outperforming traditional empirical methods. Consolidated metrics demonstrate that GNNs achieve ROC-AUC values exceeding 0.92 in classifying active compounds, drastically reducing false-positive rates. Furthermore, generative models have proven to compress the hit-to-lead timeline by up to 80%. Finally, current computational bottlenecks, such as training data bias and synthetic accessibility, are discussed alongside the future perspectives of autonomous laboratories and quantum machine learning.

Fields of Study & Subjects

Computational Drug Discovery Methods Machine Learning in Materials Science Synthesis and Characterization of Heterocyclic Compounds

Publication details

Journal / source
ChemRxiv
Publication date
25 March 2026
Language
EN
Document type
Preprint