Identifying Genetic Load By Whole Genome and Transcriptome Sequencing of Cassava Breeding Lines

Cassava (Manihot esculenta) is a major staple crop of 500 million people across the globe. Cassava is mostly vegetatively propagated. Large amount of genetic load (deleterious mutations) are accumulated due to limited recombination in this species. Significant reduction in root yield (~63%) was reported in first generation of cassava plants due to genetic load. Present study aimed at the identification of variants using whole genome sequencing of 130 cassava breeding lines at 20X coverage. These variants will be annotated with key attributes including evolutionary constraints, functional loci, etc, supplemented with RNA-Sequencing of 6 important clones and each with 5 different tissues. First and second generation of eight cassava mapping populations will be used to characterize the nature of these deleterious mutations. Using these variants and attributes, we will develop machine learning models to predict the effects of deleterious mutations on different traits in subsequent generations, primarily focusing on root yield. The models will be evaluated on 8,000 of cassava breeding lines under NextGen Cassava project. Standard genomic selection methods and/or CRISPR will be considered to purge genetic load in Cassava.