Employing AI-guided approaches to discover and engineer plant disease resistance proteins

点击翻译按钮获取中文摘要

Plant diseases threaten our food supply. Pathogens can reduce crop yields by 30%, costinghundreds of billions of dollars. Losses could feed billions. Global trade, climate change, and plantpathogens' tendency to reduce disease resistance have increased plant disease outbreaks.Population is expected to reach 9 billion by 2050, and food production must increase by 70% toprevent global hunger.Here we propose an AI-guided synthetic biology approach to discover and engineer plant immunereceptors with new disease resistance gene specificities. To this end we formed a partnership withResurrect bio, to implement an AI-guided pipeline combined with our expertise in plantimmunology. We will first establish an automated machine learning pipeline to generate digitallibraries of ligand-receptor complexes from plant-pathogen genomes. Next, ligands of interest willbe screened using AF2-multimer to identify candidate receptors, which will be subsequentlyvalidated through genetic and biochemical approaches. We will then transform individualreceptors into plant species that lack them, as a proof of concept for engineering diseaseresistance. However, some pathogens encode virulent alleles of the ligands of interest, enablingthem to avoid immune recognition. We will identify such ligands from genomes of the virulentpathogen strains through genomic surveys and make necessary mutations in the correspondingreceptor interfaces to expand their resistance spectrum. Our business partners, Resurrect Bio,specialized to discover and amend defeated disease resistance traits will then employ genome-editing approaches to introduce these traits in the crops of interest