Genotyping of DNA pools identifies untapped landraces and genomic regions to develop next‐generation varieties

Abstract

Summary Landraces, that is, traditional varieties, have a large diversity that is underexploited in modern breeding. A novel DNA pooling strategy was implemented to identify promising landraces and genomic regions to enlarge the genetic diversity of modern varieties. As proof of concept, DNA pools from 156 American and European maize landraces representing 2340 individuals were genotyped with an SNP array to assess their genome‐wide diversity. They were compared to elite cultivars produced across the 20th century, represented by 327 inbred lines. Detection of selective footprints between landraces of different geographic origin identified genes involved in environmental adaptation (flowering times, growth) and tolerance to abiotic and biotic stress (drought, cold, salinity). Promising landraces were identified by developing two novel indicators that estimate their contribution to the genome of inbred lines: (i) a modified Roger's distance standardized by gene diversity and (ii) the assignation of lines to landraces using supervised analysis. It showed that most landraces do not have closely related lines and that only 10 landraces, including famous landraces as Reid's Yellow Dent, Lancaster Surecrop and Lacaune, cumulated half of the total contribution to inbred lines. Comparison of ancestral lines directly derived from landraces with lines from more advanced breeding cycles showed a decrease in the number of landraces with a large contribution. New inbred lines derived from landraces with limited contributions enriched more the haplotype diversity of reference inbred lines than those with a high contribution. Our approach opens an avenue for the identification of promising landraces for pre‐breeding.

Genomic, climatic, and cultural diversity of maize landraces from the Himalayan Kingdom of Bhutan

“…The objective of this strategy was to increase the chances to capture all the diversity available at the landraces level; the downside being an increased complexity in the effort to accurately detect alleles in bulked samples. Effective methods for SNP calling in pooled samples have been developed on genotyping array platforms (Arca et al, 2021), enabling the characterization of large pools of maize landraces (Arca et al, 2023). In our case, we used sequencing read counts for each allele at a given genomic position factoring in base quality, mapping quality, and read depth to weigh on the representativeness of each marker using VarScan (Koboldt et al, 2009).…”

Genetic and Phenotypic Evaluation of European Maize Landraces as a Tool for Conservation and Valorization of Agrobiodiversity

“…We estimated the genetic distance between all landraces using modified Roger’s distance (MRD) [ 31 ] with ad hoc scripts, and fixation index (F ST ) based on the allelic frequencies of 23,412 SNPs using R package HFst [ 32 ]. To decipher the structure of genetic diversity within our panel of 626 landraces and their relations with worldwide diversity previously described in [ 33 ], we used three approaches: A distance-based approach using MRDs between the 626 landraces to perform principal coordinate analysis (PCoA) [ 34 ], as well as hierarchical clustering. We applied either Ward or neighbor-joining algorithms using the “hc” and “bionj” functions of “ape” R package v 5.0 [ 35 ], respectively; A Bayesian multi-locus approach, implemented in the ADMIXTURE v1.3.0 software, to assign probabilistically each landrace to K ancestral populations assumed to be in Hardy–Weinberg equilibrium [ 36 ].…”

“…Different methods were used to identify the most appropriate number of ancestral populations (K): cross-validation error or difference between successive cross-validations [ 36 ] and graphical methods [ 37 ]. Since ADMIXTURE requires multi-locus genotypes of individual plants, we simulated the genotype of five individuals for each population for a subset of 2500 independent SNPs to avoid artifacts of linkage disequilibrium (See Method S5 in [ 33 ] for more details). A linear penalized regression approach, to quantitatively assign each of the 626 landraces to seven genetic groups established for 156 landraces representing European and American diversity [ 33 ].…”

“…Since ADMIXTURE requires multi-locus genotypes of individual plants, we simulated the genotype of five individuals for each population for a subset of 2500 independent SNPs to avoid artifacts of linkage disequilibrium (See Method S5 in [ 33 ] for more details). A linear penalized regression approach, to quantitatively assign each of the 626 landraces to seven genetic groups established for 156 landraces representing European and American diversity [ 33 ]. Allelic frequencies at 23,412 SNPs of each of the seven genetic groups were estimated considering landraces with a membership superior to 0.6.…”

Fruit Agronomic and Quality Traits of Tomato F1 Hybrids Derived from Traditional Varieties

“…Currently, the nutritional and organoleptic quality of fresh tomatoes is a vital parameter for growers and tomato breeders due to the rising consumer interest in healthy and flavourful food [5,22,23]. Landraces could serve as a workable solution, as they may possess beneficial allelic combinations that are absent in modern varieties [24,25]. Although these traditional types often possess exceptional organoleptic and nutritional characteristics, their agronomic performance in terms of yield or resistance to the prevailing biotic stresses is often unsatisfactory, rendering them inappropriate for widespread commercial cultivation under most circumstances [26].…”