Drug Repurposing and De Novo Novel Monoclonal Antibody Design for Chikungunya Virus: A Two-pronged Bioinformatics Analysis

Abstract

Introduction

Chikungunya, a WHO-designated neglected tropical disease, has no existing antiviral treatment. Climate-driven vector expansion and worsening endemicity in tropical regions are escalating its global threat, particularly for resource-limited communities. Two viral proteins are compelling therapeutic targets: non-structural protein 2 (nsP2), essential for replication and immune evasion, and the E2 surface glycoprotein, the primary neutralization target.

This two-pronged AI-driven approach aims to identify repurposable drugs for immediate use targeting the Chikungunya virus (CHIKV) nsP2 and to design novel antibodies targeting the E2 surface protein.

Method 

The whole DrugBank library (14,612 ligands) was screened against nsP2 using the AI model Chai-1. By jointly modelling within a unified diffusion framework, Chai1 generated high-fidelity 3D nsP2-ligand binding conformations. Drug candidates were then prioritized by confidence score and physicochemical properties.

1000 antibody sequences against CHIKV E2 were generated using the fine-tuned MAGE protein language model. After humanness filtering and antibody-antigen structural modelling, top 10 candidates were assessed for antigen-binding affinity via enzyme-linked immunosorbent assay (ELISA).

Results

Nine FDA-approved drugs were identified to have high binding confidence scores (>0.80) to nsP2. The top three candidates are the acyclic guanosine analogs penciclovir, acyclovir, and ganciclovir, currently approved for herpesviruses. Five investigational drugs with a confidence level > 0.85 were also identified.

One of the top ten novel monoclonal antibodies exhibited strong dose-dependent binding to E2, with the highest OD readings at lower dilutions.

Conclusion

This dual-pronged approach successfully identified 14 existing drugs, including nine FDA-approved drugs and five investigational drugs, along with one high-potency novel monoclonal antibody as promising antiviral candidates against CHIKV. While further in vitro, in vivo, and clinical trials are essential to validate our findings, these results provide a pipeline for both immediate drug repurposing and long-term immunotherapeutic development to address the need for effective antivirals in endemic regions.

Keywords

Chikungunya, drug repurposing, monoclonal antibodies, AI models, bioinformatics