What question did this study set out to answer?

The aim is to review in silico methods for developing selective inhibitors targeting human carbonic anhydrases.

May 3, 2026

Computational Approaches and Structure-Based Drug Design of CAIs.

Key Points

The aim is to review in silico methods for developing selective inhibitors targeting human carbonic anhydrases.
Overview of computational methods including pharmacophore modeling, virtual screening, and molecular dynamics.
Case studies illustrating structure-based and ligand-based approaches for selective inhibitor discovery.
Discussion on AI and machine learning frameworks applied to biochemical datasets for enhanced predictive modeling.
Demonstrated successful identification of selective hCA inhibitors through computational strategies.
Highlighted the importance of structure-function relationships in rational ligand design.
Integrated AI tools showed potential in increasing accuracy in predicting inhibitor potency and selectivity.

Abstract

Human carbonic anhydrases (hCAs) are emerging as increasingly relevant therapeutic targets due to their involvement in a broad spectrum of pathological conditions, ranging from glaucoma and epilepsy to cancer, neuroinflammation, metabolic disorders, and obesity. The coexistence of 15 isoforms with distinct catalytic activities and tissue distributions makes selectivity a central challenge in the development of CA inhibitors (CAIs). Recent advances in computational chemistry, artificial intelligence (AI), machine learning (ML), and Structure-Based Drug Design (SBDD) have significantly accelerated the identification of new chemotypes and selective modulators.This chapter provides a comprehensive overview of in silico strategies applied to CAIs, including pharmacophore modeling, Virtual Screening, docking, molecular dynamics (MD), MM-GBSA, and AI-driven predictive modeling. Case studies illustrate how Ligand-Based and Structure-Based approaches have led to the discovery of selective inhibitors for hCAs. A special emphasis is placed on structure-function relationships, including peculiar active-site features responsible for isoform differentiation and exploitable for rational ligand design.Furthermore, the chapter highlights recent ML- and DL-based frameworks trained on large biochemical datasets, capable of predicting potency and selectivity with high accuracy. Notably, the integration of explainable AI tools and experimental validation demonstrates the potential of computational pipelines in guiding hit identification and optimization.Overall, this chapter underscores the increasingly synergistic role of computational methodologies and AI in the development of selective CAIs.

Mark Helpful

Bookmark

Relay