Genotype concordance and trait mapping efficacy comparing data from the Equine 670 K SNP array with whole genome sequence in 21 horses.

Overview

• This study compares the accuracy and effectiveness of two genomic technologies—the Equine 670K SNP array and whole genome sequencing (WGS)—in identifying genetic variants and mapping traits in horses.
• The research focuses on concordance in genotype calls and the ability to map a specific coat color trait in 21 horses genotyped by both methods.

Background

• Genomic technologies like SNP arrays and whole genome sequencing are widely used for studying horse genetics.
• SNP arrays genotype a set fixed number of genetic variants across the genome, in this case 670,000 SNPs known from previous studies.
• WGS sequences the entire genome allowing for detection of all possible variants, but data quality depends on sequencing depth (coverage).

Methods

• 21 horses were both genotyped using the Equine 670K SNP array and sequenced at approximately either 12× or 30× coverage depth.
• Genotype calls (SNP variants) from the SNP array and WGS data were compared to assess concordance—how often they called the same genotype at given positions.
• Discordant and no-call positions were further examined to understand causes, including whether variants were insertions/deletions (indels) or multiallelic polymorphisms in the WGS data.
• Trait-mapping effectiveness was evaluated by association analyses on a known simple trait—chestnut coat color—which is controlled by a major gene.

Key Findings

• Higher WGS coverage (30×) resulted in fewer discordant genotype calls compared to lower coverage (~12×), indicating higher sequencing depth improves accuracy.
• Some positions commonly no-called or discordant were due to complex variants like indels or multiallelic sites, which are more challenging to genotype accurately using SNP arrays designed for biallelic SNPs.
• Both the SNP array and WGS datasets could detect the chestnut coat color association signal prominently at the expected genomic locus.
• However, neither method reached statistical significance after strict Bonferroni correction, likely due to the small sample size (n=21) limiting power.

Implications

• The study demonstrates that while both platforms can be used for SNP genotyping and trait mapping, WGS provides more comprehensive variant detection especially at higher coverage depths.
• SNP arrays remain useful and produce reliable genotype calls but may miss complex variant types and novel alleles not included in the array design.
• For small sample sizes, neither platform is likely to achieve stringent genome-wide significance in mapping studies, highlighting the need for larger cohorts.
• This data helps researchers select the appropriate genomic technology based on trade-offs between cost, desired variant detection, data quality, and study design.

Conclusion

• Genotype concordance between the Equine 670K SNP array and WGS depends on WGS sequencing depth, with higher coverage minimizing discordances.
• The ability to map a simple trait is comparable but limited by sample size, regardless of technology used.
• These results provide valuable guidance for equine genomic research planning and interpretation when choosing between SNP arrays and WGS platforms.