Computational strategies for elucidating plant disease resistance proteins

Abstract

Plant disease resistance proteins (R-proteins) play a crucial role in initiating immune responses by recognizing pathogen-derived signals and triggering downstream defense mechanisms. This review presents an in-depth evaluation of both bioinformatics approaches and advance computational techniques for the identification and characterization of R-proteins across diverse plant species. Particular emphasis is placed on the transformative impact of machine learning (ML) and deep learning (DL) in R-gene discovery and classification. ML algorithms facilitate advanced modeling of complex sequence features and classification tasks, surpassing the limitations of conventional similarity-based methods. Moreover, deep learning architectures such as Convolutional Neural Networks (CNNs), Multi-Layer Perceptrons (MLPs) and Recurrent Neural Networks (RNNs) models have proven highly effective in capturing hierarchical and contextual information from biological sequences, thereby improving prediction accuracy and enhancing model generalization. The review also surveys several key curated databases, including PRGdb, the NBS-LRR Receptor database, SolRgene, RiceMetaSysB, LDRGDb, PlantNLRatlas, and RefPlantNLR, which collectively support robust annotation and comparative analysis of R-genes across species. The integration of machine learning and deep learning models with these databases accelerates the identification of novel R-proteins and deepens our understanding of plant immunity. This synergy provides powerful tools for breeding disease-resistant crops and supports the broader goals of sustainable and resilient agriculture.