An integrated framework reinstating the environmental dimension for GWAS and genomic selection in crops

Abstract

Identifying mechanisms and pathways involved in gene-environment interplay and phenotypic plasticity is a long-standing challenge. It is highly desirable to establish an integrated framework with an environmental dimension for complex trait dissection and prediction. A critical step is to identify an environmental index that is both biologically relevant and estimable for new environments. With extensive field-observed complex traits, environmental profiles, and genome-wide single nucleotide polymorphisms for three major crops (maize, wheat, and oat), we demonstrated that identifying such an environmental index (i.e., a combination of environmental parameter and growth window) enables genome-wide association studies and genomic selection of complex traits to be conducted with an explicit environmental dimension. Interestingly, genes identified for two reaction-norm parameters (i.e., intercept and slope) derived from flowering time values along the environmental index were less colocalized for a diverse maize panel than for wheat and oat breeding panels, agreeing with the different diversity levels and genetic constitutions of the panels. In addition, we showcased the usefulness of this framework for systematically forecasting the performance of diverse germplasm panels in new environments. This general framework and the companion CERIS-JGRA analytical package should facilitate biologically informed dissection of complex traits, enhanced performance prediction in breeding for future climates, and coordinated efforts to enrich our understanding of mechanisms underlying phenotypic variation.

Designing Future Crops: Genomics-Assisted Breeding Comes of Age

“…Similar to the proposition of 'landscape genomics' in ecology, environmental GWAS in crops uses the climatic conditions as phenotype to identify single nucleotide polymorphisms (SNPs) associated with the accession's environment of origin. To this end, a more recent study by Li et al [20] provided an integrated framework to conduct GWAS and GS in crops with an environmental dimension. Finding patterns in environmental index and associating these patterns with changes in underlying genomic determinants have great implications for understanding complex traits in plants and their prediction for future climates.…”

Can We Harness “Enviromics” to Accelerate Crop Improvement by Integrating Breeding and Agronomy?

“…There have been many calls for enhanced attention to environmental characterization to accelerate crop improvement. Plant breeders have long sought environmental definitions and covariates to assist interpretation of plant responses and the GxE interactions detected in METs and to understand their relevance for the on-farm TPE (Finlay and Wilkinson, 1963;Allard and Bradshaw, 1964;Baker, 1988;Blum, 1988;Cooper and Hammer, 1996;Boer et al, 2007;Heslot et al, 2014;Jarquín et al, 2014;Pauli et al, 2016;Xu, 2016;Ly et al, 2018;Bustos-Korts et al, 2019Millet et al, 2019;Costa-Neto et al, 2020Porker et al, 2020;Crossa et al, 2021;Li et al, 2021;Resende et al, 2021;Smith et al, 2021). The role of water availability and impact of drought on crop yield and investigations to determine the traits contributing to crop productivity under drought conditions have received significant attention from breeders (e.g., Blum, 1988;Fukai and Cooper, 1995;Campos et al, 2004;Bänziger et al, 2006;Ribaut, 2006;Messina et al, 2011Messina et al, , 2018Cooper et al, 2014a), agronomists (French and Schultz, 1984;Sadras and Angus, 2006;Kirkegaard and Hunt, 2010;Van Ittersum et al, 2013;Hunt et al, 2021), and physiologists (Richards and Passioura, 1989;Ludlow and Muchow, 1990;…”

Predicting phenotypes from genetic, environment, management, and historical data using CNNs

“…Prediction of phenotypes from a combination of environmental (E), genetic (G), and humanimposed (often referred to as management(M)) conditions has been a long standing challenge in biology and related fields (Messina et al 2009(Messina et al , 2018Technow et al 2015;Cooper et al 2016Cooper et al , 2021Varshney et al 2017;Washburn et al 2020;Jarquin et al 2021;Li et al 2021).…”

“…Geneticists and breeders have successfully used statistical models, primarily Best Linear Unbiased Prediction (BLUP) and genomic BLUP (GBLUP), for decades (Henderson 1975;Meuwissen et al 2001;Gaffney et al 2015). These methods have focused on prediction within the G space (sometimes called genomic prediction (GP)) but can be extended to incorporate information from E and/or M (Jarquín et al 2014;Li et al 2021). The downside to these methods is that they are not mechanistic in nature and require extensive feature engineering and data complexity reduction in order to avoid overfitting, limiting their interpretation, and perhaps accuracy.…”