Dual-Genome Prediction Model for Black-Pod-Resistant Cacao

Integrates Host- and Pathogen-Genome Data in a Unified Pipeline to Evaluate Prediction Accuracy and Accelerate Breeding Decisions

This dual-genome prediction model for black-pod-resistant cacao integrates host- and pathogen-genome data with phenotypic data, such as disease resistance ratings, in a unified pipeline, allowing for a detailed evaluation of prediction accuracy and accelerating breeding decisions. Black pod disease is a major threat to cacao production, significantly reducing yields and bean quality in growing regions worldwide. Phytophthora megakarya is one of the pathogens that cause black pod rot. It is endemic to West Africa, causing severe damage in the main cacao-producing countries. Current resistance-screening methods are slow and resource-intensive, often yielding inconsistent results across regions. The global cocoa bean market was valued at $13.54 billion in 2023 and is projected to reach $23.54 billion by 2030. Rising consumer demand for cocoa-rich products, driven by health benefits, highlights the market need for tools that enable breeders to identify resistant germplasm quickly and accurately. Sustainable cocoa production now hinges on dependable, disease-resistant varieties, so an accurate approach for predicting resistance across diverse cacao-pathogen genotype combinations is necessary.

University of Florida researchers, in collaboration with Mars Inc. and USDA partners, developed this dual-genome prediction platform for black-pod-resistant cacao, which integrates host- and pathogen-genome data in a unified pipeline to evaluate the accuracy of the predictions and accelerate breeding decisions. By linking cacao and Phytophthora?megakarya genomes, the platform captures host-by-pathogen interaction effects, delivering higher predictive power than single-genome methods and enabling rapid evaluation. This proof-of-concept system may be applied to other cacao diseases and adapted for use with other crops, providing a powerful addition to integrated pest management and plant genetics programs worldwide.

Application

This dual-genome prediction model enables breeders to accurately identify cacao varieties resistant to black pod disease, facilitating targeted deployment of resilient crops in vulnerable production regions

Advantages

• Integrates host and pathogen genomic data, improving resistance-screen accuracy
• Cleans genomic data, reducing noise and error
• Enables rapid in-silico screening of thousands of germplasm entries, accelerating breeding cycles
• Predicts resistance pre-deployment, targeting high-risk regions for resilient yields

Technology

This dual-genome prediction model for black-pod-resistant cacao integrates host- and pathogen-genome data in a unified pipeline. The script first cleans the raw DNA data from both cacao trees and Phytophthora?megakarya. It then merges the two genomic inputs with phenotypic data, including disease severity scores and resistance ratings, to build a prediction model and predict the extent of disease damage on cacao pods. By generating accurate resistance predictions in minutes, the platform lets breeders screen thousands of cacao germplasm candidates without field trials, enabling the strategic introduction of resilient cacao varieties to regions most affected by black pod disease. While developed for cacao, the system’s flexible design allows it to be applied to a wide range of crops and plant diseases, offering a scalable solution for improving agricultural resilience worldwide.