Genomic Prediction of Agronomic Traits in Common Bean (Phaseolus vulgaris L.) Under Environmental Stress

Abstract

developed models allowed us to predict genotypic performance under different environmental stresses. This will be a key factor in the development of common bean varieties adapted to future challenging conditions.

Genetic Architecture and Genomic Prediction of Cooking Time in Common Bean (Phaseolus vulgaris L.)

“…This population contained 636 F 4:6 RILs generated by the SSD method. (3) The “Vivero Equipo Frijol” (VEF) panel, previously described by Keller et al (2020) is composed of 380 Andean breeding lines. (4) The “Mesoamerican Introgression Panel” (MIP) consists of 222 breeding lines of the Mesoamerican gene-pool.…”

“…A subset of 223 MAGIC lines and its eight founders were phenotyped for cooking time. The VEF panel was planted in 2017 with an alpha-lattice incomplete block experimental design of three replicates as described by Keller et al (2020) . The MIP panel was grown in 2018 following an alpha-lattice incomplete block design with three replicates.…”

“…Genotyping of the MAGIC population and its founders (629 RIL + 8 lines), the VEF panel (330 lines), the MIP panel (210 lines), and each founder genotype of the DxG population was obtained using the Ape KI-based genotyping-by-sequencing (GBS) protocol ( Elshire et al, 2011 ), as previously described ( Perea et al, 2016 ; Gil et al, 2019 ; Keller et al, 2020 ; Diaz et al, 2020 ). Briefly, DNA was extracted using the urea buffer-based DNA extraction midi prep protocol.…”

“…The mapping and variant calling processes for GBS reads is described in detail by Perea et al (2016) , Gil et al (2019) , and Keller et al (2020) . In brief, the GBS reads were demultiplexed using NGSEP (v3.3.2) ( Tello et al, 2019 ).…”

“…A single-trait assessment of prediction ability was performed for each population individually using the R package BGLR (v1.0.8) ( Pérez and De Los Campos, 2014 ), with 10,000 iterations, using the first 2,000 for burn-in and default parameters. In each case, 70% of the individuals in the population were assigned to the training set, while the remaining 30% were assigned to the validation set, following the results from Keller et al (2020) . This random assignment was repeated 100 times for each population.…”