Multi-trait random regression models increase genomic prediction accuracy for a temporal physiological trait derived from high-throughput phenotyping

Abstract

29Random regression models (RRM) are used extensively for genomic inference and prediction 30 of time-valued traits in animal breeding, but only recently have been used in plant systems. 31 High-throughput phenotyping (HTP) platforms provide a powerful means to collect high- 32 dimensional phenotypes throughout the growing season for large populations. However, to 33 date, selection of an appropriate statistical genomic framework to integrate multiple temporal 34 traits for genomic prediction in plants remains unexplored. Here, we demonstrate the utility 35 of a multi-trait RRM (MT-RRM) for genomic prediction of daily water usage (WU) in rice 36 (Oryza sativa) through joint modeling with shoot biomass (projected shoot area, PSA). 37Three hundred and fifty-seven accessions were phenotyped daily for WU and PSA over 20 38 days using a greenhouse-based HTP platform. MT-RRMs that modeled additive genetic 39 and permanent environmental effects for both traits using quadratic Legendre polynomials 40 were used to assess genomic correlations between traits and genomic prediction for WU. 41 Predictive abilities of the MT-RRMs were assessed using two cross-validation (CV) scenarios. 42 The first scenario was designed to predict genetic values for WU at all time points for a set 43 of accessions with unobserved WU. The second scenario was designed to forecast future 44 genetic values for WU for a panel of known accessions with records for WU at earlier time 45 periods. In each scenario we evaluated two MT-RRMs in which PSA records were absent or 46 available for time points in the testing population. Moderate to strong genomic correlations 47 between WU and PSA were observed across the days of imaging (0.38-0.80). In both CV 48 scenarios, MT-RRMs showed better predictive abilities compared to single-trait RRM, and 49 prediction accuracies were greatly improved when PSA records were available for the testing 50 population. In summary, these frameworks provide an effective approach to predict temporal 51 physiological traits that are difficult or expensive to quantify in large populations. 52 3 Background 53High-throughput phenotyping (HTP) is an innovative tool in plant breeding. HTP pro-54 vides precise and non-destructive estimation of multiple complex traits that describe growth 55 and development (e.g., height, biomass, and flowering time) or environmental responses 56 (e.g., chlorophyll fluorescence, canopy temperature, and water content) using non-destructive 57 image-based phenotyping (Araus et al., 2018; Morota et al., 2019). These HTP data mitigate 58 extensive costs associated with manual phenotyping, and can be used to better capture the 59 plant's phenome. In the context of plant breeding and genetics, these data can be used to 60 improve the prediction of breeding values for a target trait of interest, thereby improving 61 the accuracy of selection, as well as provide insights into how secondary traits influence a 62 trait of interest (Araus et al., 2018; Morota et al., 2019; Voss-Fels et ...

High-Throughput Phenotyping and Random Regression Models Reveal Temporal Genetic Control of Soybean Biomass Production

“…Under the field conditions of multienvironment trials, as in our study, the genetic contribution to the observed phenotypes is both variable and reduced due to environmental fluctuations. Regarding genetic correlation of phenotypes across days (Figure 4), Campbell et al (2018) and Baba et al (2020) observed the same trend that we did, where the highest correlations were observed between adjacent time points.…”

“…The genetic correlation between longitudinal soybean AGB and grain yield is currently being investigated. Given HTPP's power to simultaneously collect multiple temporal traits, multiple-trait RRM may be powerful tools for joint genomic prediction of multiple longitudinal traits (Oliveira et al, 2016;Baba et al, 2020;Moreira et al, 2020). Therefore, RRM and HTPP have a great potential to accelerate the rate of genetic gain in soybean breeding programs.…”

Application of Genomics and Phenomics in Plant Breeding for Climate Resilience

“…Genomics-Assisted Breeding (GAB) methodologies encompass an array of stages including characterization of germplasm collections; mapping population development; genomic region identification through association or genetic mapping; and application of molecular markers [24,25]. GAB rose to popularity in the field of plant breeding in the past few years partly due to the presence of low-cost NGS and highthroughput genotyping (HTPG) technologies [25] which allowed the rapid identification of superior genes implicated in climate resilience [26,3]. The fight against climate change needs immediate responses to counter the associated negative effects on crop productivity.…”