LUCID: An Integrative Approach for Target Discovery and dsRNA Design in Plant Fungal Pathogens

Abstract

Phytopathogenic fungi pose an escalating threat to global food security and ecosystem stability, as resistance and environmental concerns diminish the effectiveness of conventional fungicides. Double-stranded RNA (dsRNA)-based fungicides offer a species-specific, eco-friendly alternative. We introduce LUCID (Locating Uncovered, conserved, and Indispensable for pathogenicity Determinants), a computational pipeline that accelerates the development of RNAi-based biofungicides by integrating target identification with dsRNA design and off-target prediction. LUCID employs a dual-branch strategy to identify both Conserved Essential Proteins (CEPs) and Conserved Non-Annotated Proteins (CNAPs), leveraging transcriptomic data and comparative genomics across diverse fungal species. Validation in Botrytis cinerea demonstrated high efficacy, with 67% of proposed targets successfully silenced and an average silencing efficiency of 96%. Additionally, coupling LUCID with advanced protein language models (PLMs) revealed a novel pathogenicity determinant in B. cinerea: a putative mediator complex protein. LUCID offers a scalable, species-agnostic framework for designing sustainable fungicides, enabling rapid, targeted control of fungal diseases with minimal ecological impact.