Genotype imputation from low-coverage WGS using haplotype reference panels in cultivated strawberry

Area of Science:

• Genomics
• Plant Breeding
• Bioinformatics

Background:

• High-throughput sequencing-based genotyping is crucial for strawberry (Fragaria × ananassa) breeding programs.
• Whole genome sequencing (WGS) offers high SNP density but is costly for large-scale applications.
• Genotype dosage imputation from low-coverage data using a reference panel is a cost-effective alternative, but requires optimization for allo-octoploid strawberry.

Purpose of the Study:

• To construct a haplotype reference panel for strawberry.
• To explore the utility of this panel for genotype dosage imputation of low-coverage (1×) sequencing data.
• To optimize imputation strategies for allo-octoploid strawberry.

Main Methods:

• Combined high sequencing depth (>15×) with variant filtering (AAB, LD, MER) to reduce genotyping errors.
• Employed statistical phasing (SHAPEIT5) achieving a mean switch error rate of 0.9%.
• Utilized GLIMPSE2 for imputation of downsampled (1×) samples using reference panels of varying size and composition.

Main Results:

• Achieved accurate statistical phasing with 50% of the genome in haplotype blocks of at least 654 kb.
• Imputation accuracy varied with reference panel size and genetic diversity, yielding concordance rates from 0.87 to 0.98.
• Demonstrated that panel size and genetic diversity significantly influence imputation accuracy.

Conclusions:

• A large, genetically diverse haplotype reference panel enhances genotype dosage imputation from low-coverage sequencing data.
• High accuracy imputation is achievable with limited resources, offering a cost-efficient alternative to SNP arrays for WGS-based genotyping.
• The developed strategy is applicable to other crops needing dense genotyping with limited resources, suggesting ~70 samples at ≥25× depth suffice for panel construction.