Recommendations for Choosing the Genotyping Method and Best Practices for Quality Control in Crop Genome-Wide Association Studies

Abstract

High-throughput genotyping boosts genome-wide association studies (GWAS) in crop species, leading to the identification of single-nucleotide polymorphisms (SNPs) associated with economically important traits. Choosing a cost-effective genotyping method for crop GWAS requires careful examination of several aspects, namely, the purpose and the scale of the study, crop-specific genomic features, and technical and economic matters associated with each genotyping option. Once genotypic data have been obtained, quality control (QC) procedures must be applied to avoid bias and false signals in genotype-phenotype association tests. QC for human GWAS has been extensively reviewed; however, QC for crop GWAS may require different actions, depending on the GWAS population type. Here, we review most popular genotyping methods based on next-generation sequencing (NGS) and array hybridization, and report observations that should guide the investigator in the choice of the genotyping method for crop GWAS. We provide recommendations to perform QC in crop species, and deliver an overview of bioinformatics tools that can be used to accomplish all needed tasks. Overall, this work aims to provide guidelines to harmonize those procedures leading to SNP datasets ready for crop GWAS.

Keywords: GWAS; bioinformatics tools; crops; genotyping; quality control.

FIGURE 1

Overview of quality control procedures for crop GWAS. These include: filtering steps that are common to any GWAS experiment; filtering steps depending on the GWAS population structure (homozygous or heterozygous); the removal of duplicated samples; the characterization of ancestral relationships, starting from a SNP dataset pruned for markers in linkage disequilibrium.

FIGURE 2

Frequency distribution analysis to define filtering solutions for (A) SNP call rate; (B) genotype call rate; (C) SNP inbreeding coefficient (F_IT); (D) SNP proportion of heterozygosity. Dashed lines indicate possible filtering thresholds, based on classes occurring at suspiciously low (A,B) or high (D) frequency, and distribution gaps (C). Genotypic data used to build histograms are all relative to published genotyping-by-sequencing experiments, carried out in the self-pollinated crops Cicer arietinum L. (Pavan et al., 2017, A,C) and Lens culinaris Medik (Pavan et al., 2019, B), and the open-pollinated crop Cynara cardunculus L. (Pavan et al., 2018, D).