Predicting herbicide resistance through phylogenomics

Overview

This solution offers a predictive framework for identifying metabolic herbicide resistance in plants before it manifests in the field. By combining phylogenomics, metabolomics, and cheminformatics, the approach builds a "digital ADME twin" of a plant species using data from its close botanical relatives. The framework models the evolution of conjugating enzymes responsible for non-target-site resistance and predicts which plant genotypes are likely to inactivate specific herbicides through enzymatic modification. This enables agrochemical companies, seed producers, and crop protection researchers to anticipate resistance emergence, prioritize herbicide candidates less prone to metabolic breakdown, and design new molecules that resist common detoxification pathways.

Technical specifications

Core methodology:

• Phylogenomics modeling of gene family expansion across plant families to infer diversity of conjugating enzymes (transferases)
• Metabolomics integration to map secondary metabolite diversity and scaffold decoration patterns across related species
• Cheminformatics analysis to evaluate chemical similarity between herbicides and native plant compounds
• LC-MS validation to track herbicide conjugates formed when challenged plants are exposed to chemical treatments
• Heterologous expression and biochemical testing of candidate enzymes to confirm predicted conjugation activity

Key features:

• Predicts resistance based on evolutionary relationships rather than waiting for field failure
• Applicable across Monocots (including Pooideae grasses such as Alopecurus and Triticum) and Dicots (including Brassicaceae such as Arabidopsis and Brassica)
• Compares predictions against 269 known herbicide-resistant species for validation
• Provides enzyme-level insight into which herbicide molecular positions are vulnerable to attack

Technology readiness level

The technology is at an early-to-mid validation stage. Preliminary analysis has been completed using untargeted metabolomics data from 14 species across multiple plant families, combined with their transcriptomes, successfully relating gene family expansion to chemical diversity emergence. Candidate enzymes have been ranked by predicted conjugation activity. Cloning, expression, and biochemical testing of top candidates are scheduled for the first half of 2024. Future validation will scale the model across 767 species with available metabolomes, apply predictions to over 35 Pooideae genomes and over 70 Brassicaceae genomes, and validate predictions through seed challenge experiments, mass spectrometry analysis, and enzyme purification studies.