Utilizing evolutionary conservation to detect deleterious mutations and improve genomic prediction in cassava

Area of Science:

• Genetics
• Plant Breeding
• Bioinformatics

Background:

• Cassava (Manihot esculenta) is a vital global food source, primarily propagated clonally, leading to inbreeding depression due to accumulated deleterious mutations.
• Understanding and characterizing these mutations is crucial for improving cassava breeding and crop performance.

Purpose of the Study:

• To locate and characterize deleterious mutations across the cassava genome.
• To measure selection pressure and estimate evolutionary constraint.
• To improve genomic evaluations of cassava plant performance using genomic prediction.

Main Methods:

• Aligned 52 related Euphorbiaceae species to analyze genetic conservation at single base-pair resolution.
• Utilized protein structure models, amino acid impact, and evolutionary conservation to estimate constraint.
• Employed multi-kernel GBLUP (Genomic Best Linear Unbiased Prediction) for genomic prediction simulations.

Main Results:

• Incorporating functional variants significantly increased prediction accuracy for traits influenced by fewer than 100 quantitative trait loci (QTL).
• Deleterious mutations and functional weights derived from evolutionary conservation improved genomic prediction accuracy, with variations based on trait and prediction method.

Conclusions:

• Evolutionary information can effectively track functional genomic variation to enhance whole-genome trait prediction in cassava.
• Further improvements in genotype accuracy and deleterious mutation assessment will refine genomic evaluations for cassava clones.