Machine learning, bioinformatics analysis and chemical screening streamline target validation by identifying RNA...

Plant viruses inflict annual economic losses exceeding $30 billion and pose a significant threat to global food security. Discovering reliable antiviral targets is therefore essential to developing effective agents that can substantially reduce these agricultural losses. Despite ongoing efforts, the rapid discovery of such antiviral targets remains a significant challenge.

This study employs an accelerated framework that integrates a machine learning-driven (MLD) approach with bioinformatics and in silico chemical screening to rapidly predict essential plant viral genes and validate novel antiviral targets.

A MLD web-based prediction tool, Vgep, was developed to predict the viral essential gene. Subsequent phylogenetic, structural, and target-likeness analyses assessed the targetability of the essential protein it encodes. Finally, in silico virtual screening, biological activity evaluation, virus morphology observation, gene expression analysis, and molecular simulations were employed to identify a chemical probe for evaluating the druggability of the essential protein.

Using the created Vgep tool, we predicted viral essential genes and identified the viral helicase as a key target in tobamoviruses. Viral helicase reveals high conservation and similarity to benchmark antiviral targets. The chemical probe Amidoca, identified through virtual screening exhibits strong binding affinity (K_d = 4.59 µM) to tobacco mosaic virus helicase. In antiviral bioassays, Amidoca outperformed Ribavirin, exhibiting EC₅₀ values of 155.85 mg/L (inactive), 244.32 mg/L (curative), and 344.59 mg/L (protective), thereby confirming the druggability of the tobamovirus helicase. Mechanistic studies revealed that Amidoca may competitively bind at the helicase's NTP-binding site. This interaction may interfere with the energy release required for unwinding viral dsRNA and lead to a dramatic reduction in helicase accumulation by nearly 95%.

This work establishes a systematic framework for the rapid discovery and validation of next-generation targets in plant virus therapies, highlighting RNA helicase as a promising antiviral candidate and advancing antiviral agent development efforts.