Alternative measures to evaluate the accuracy and bias of genomic predictions with censored records

Abstract

This study aimed to propose and compare metrics ofaccuracy and bias of genomic prediction of breeding values for traits with censored data. Genotypic andcensored-phenotypic information were simulated for four traitswith QTL heritability and polygenic heritability, respectively: C1: 0.07-0.07, C2: 0.07-0.00, C3: 0.27-0.27, and C4: 0.27-0.00. Genomic breedingvalues were predicted using the Mixed Cox and Truncated Normal models. The accuracy of the models was estimated based on the Pearson (PC), maximal (MC),and Pearson correlation for censored data (PCC) while the genomic bias was calculated via simple linear regression (SLR) and Tobit (TB). MC and PCCwere statistically superior to PC for the traitC3 with 10and 40% censored information, for 70% censorship, PCC yielded better results than MC and PC. For the other traits, the proposed measures were superior or statistically equal to the PC. The coefficients associated with the marginal effects (TB) presented estimates close to those obtained for the SLRmethod, while the coefficient related to the latent variable showed almost unchanged pattern with the increase in censorship in most cases. From a statistical point of view, the use of methodologiesfor censored data should be prioritized, even for low censoringpercentages