Exome sequence genotype imputation in globally diverse hexaploid wheat accessions

Area of Science:

• Plant genetics
• Genomics
• Bioinformatics

Background:

• Genotype imputation is crucial for cost-effective genomic analyses like GWAS and genomic prediction.
• Imputation accuracy is influenced by reference population selection, marker density, and imputation algorithms.
• Previous studies focused on low-density SNP imputation to higher-density SNP arrays.

Purpose of the Study:

• To evaluate imputation accuracy from 90K SNP chip to exome sequence in wheat.
• To investigate factors affecting imputation accuracy, including reference selection and imputation algorithms.
• To propose strategies for enhancing exome imputation accuracy in wheat.

Main Methods:

• Compared imputation accuracy from 9K to 90K SNP arrays using various reference selection and imputation algorithms (Beagle4, FImpute).
• Assessed imputation from 90K SNP array to exome sequence in wheat.
• Evaluated the impact of using exome SNPs versus 90K SNPs as the lower-density set for imputation.

Main Results:

• FImpute outperformed Beagle4 in accuracy and efficiency for SNP imputation.
• Accession-to-nearest-entry and genomic relationship-based methods were superior for reference selection.
• Imputation accuracy from 90K to exome sequence was approximately 0.71, comparable to 9K to 90K imputation.
• Accuracy improved to 0.82 when exome SNPs were used as the lower-density set instead of 90K SNPs.

Conclusions:

• Imputation from 90K SNP chip to exome sequence in wheat is moderately accurate.
• Reference population selection and imputation algorithms significantly impact accuracy.
• Using exome SNPs as the reference set substantially improves imputation accuracy, offering a promising strategy for wheat genomics.