Genome-Wide Association Studies Accelerates Genomic Selection: Implications for Cbsv Resistance in Cassava

Cassava (manihot esculenta Crantz), a key subsistence and cash crop for over 700 million people globally has its productivity threatened by the evolving range of pathogens. Cassava brown streak virus disease (CBSD) is a major biotic threat impacting root quality and marketability with a turnover of $100 million annually. Advances in genomics are revolutionizing cassava breeding opening opportunities to use tools such as genome-wide association studies (GWAS) to unveil complex genetic architecture of complex traits and genomic selection (GS) to predict plant performance before field-testing. This study demonstrates how the synergistic implementation of the two tools will greatly improve cassava breeding in the face of increasing new threats to agricultural production. GWAS identified QTLs strongly associated with CBSV resistance on chromosomes 4 and 11. The QTL in chromosome 4 co-locate with an introgression block from the wild progenitors of cassava previously identified. While a cluster of plant defense response-related genes were located near the identified QTL on chromosome 11. Predictions using a five-fold cross validation scheme had comparable accuracies across GS models though RKHS and random forest had the highest predictive values. Multi-kernel GBLUP showed that a genomic relationship matrix (GRM) constructed with markers on chromosome 11 had higher predictive accuracy for CBSD resistance scored at 3 months while a GRM from markers in chromosome 4 exhibited higher predictive ability for CBSD resistance scored at 6 months. From this study, we unraveled the genetics of CBSD resistance and demonstrated that GWAS could be used as a prioritization tool to identify markers for genomic prediction for CBSD resistance in cassava.