Regional Heritability Mapping provides insights into Dry matter (DM) Content in African white and yellow cassava populations

Okeke, U.G., Akdemir, D., Rabbi, I., Kulakow, P. & Jannink, J.L. (2018). Regional heritability mapping provides insights into Dry Matter (DM) Content in African white and yellow cassava populations. Plant Genome, 11(1), 1-18.

The HarvestPlus program for cassava (Manihot esculenta Crantz) fortifies cassava with beta-carotene by breeding for carotene-rich tubers (yellow cassava). However, a negative correlation between yellowness and dry matter (DM) content has been identified. Here, we investigated the genetic control of DM in white and yellow cassava subpopulations. We used regional heritability mapping (RHM) to associate DM to genomic segments in both subpopulations. Significant segments were subjected to candidate gene analysis and we attempted to validate candidates using prediction accuracies. The RHM procedure was validated using a simulation approach. The RHM revealed significant hits for white cassava on chromosomes 1, 4, 5, 10, 17 and 18 while hits for the yellow were on chromosome 1. Candidate gene analysis revealed genes in the carbohydrate biosynthesis pathway including the plant serine-threonine protein kinases (SnRKs), UDP-glycosyltransferases, UDP-sugar transporters, invertases, pectinases, and regulons. Validation using 1252 unique identifiers from the SnRK gene family genome-wide recovered 50% of the predictive accuracy of whole genome SNPs for DM while validation using 53 likely (extracted from literature) genes from significant segments recovered 32%. Genes including an acid invertase, a neutral/alkaline invertase and a glucose-6-phosphate isomerase were validated based on an a priori list for the cassava starch pathway and also a fructose-biphosphate aldolase from the calvin cycle pathway. The power of the RHM procedure was estimated at 47 percent when the causal QTL generated 10% of the phenotypic variance with sample size of 451. Cassava DM genetics is complex. RHM may be useful for complex traits.