Analysis of the Efficiency of Genomic Selection Models for Predicting Sheath Blight Resistance in Rice (Oryza sativa L.,)

The research was undertaken during June-October 2020 at Seethanagaram and Draksharam villages of East Godavari district, Andhra Pradesh, India with an objective to evaluate efficiency of genomic selection models involving 1545 recombinant inbred lines (RILs) derived from eleven bi-parental populations in Rice. During June-October 2020, the F7 RILs were screened in two hot spot locations. The genotyping was done with Infinium platform having 6564 SNP markers. Five models were used rrBLUP, BayesA, BayesB, BayesCPi and GBLUP to train the statistical model for calculation of marker effects and genomic estimated breeding values (GEBVs). The prediction accuracy (data fit) of training set across models ranged 0.63–0.72, lowest and highest prediction accuracies were observed with rrBLUP and GBLUP models respectively. Tenfold cross validation with different approaches, the average prediction accuracy ranged from 0.60 (rrBLUP, BayesA, BayesB and BayesCPi) –0.72 (GBLUP). BayesB and GBLUP models exhibit higher prediction accuracies compared to other models studied. The predictive ability increased dramatically with more SNPs included in analysis until 2000 markers with average prediction accuracy of 0.681, no significant improvement beyond this was observed. The results are lucrative, all in all, high prediction accuracies observed in this study suggest genomic selection as a very promising strategy while breeding for sheath blight resistance in rice to increase genetic gain.