Size and composition of haplotype reference panels impact the accuracy of imputation from low-pass sequencing in cattle

Abstract

Motivation: Low-pass sequencing followed by sequence variant genotype imputation is an alternative to the routine microarray-based genotyping in cattle. However, the impact of haplotype reference panel composition and its interplay with the coverage of low-pass whole-genome sequencing data has not been sufficiently explored in typical livestock settings where only a small number of reference samples are available. Results: Benchmarking against a microarray truth set confirms that DeepVariant is a suitable variant caller to establish bovine haplotype reference panels. We build haplotype reference panels of varying sizes and composition with DeepVariant and use GLIMPSE to impute genotypes for cattle that had between 0.01- and 4-fold genome sequence data. Same-breed haplotype reference panels perform better than equally-sized multibreed haplotype reference panels for all target sample coverages and allele frequencies. F1 scores greater than 0.9 can be achieved with 0.25-fold sequencing coverage when large breed-specific haplotype reference panels (n = 150) are used. In absence of such large same-breed haplotype panels, sequence variant genotyping accuracy can be increased either by adding non-related samples or by increasing the coverage of the low-pass sequencing data. Sequence variant genotyping from low pass sequencing is substantially less accurate when the reference panel lacks individuals from the target breed. Availability: Sequencing data are publicly available as indicated in the Supplementary Files. Scripts and workflows are available on Github: https://github.com/AnimalGenomicsETH/Low_pass_imputation. Contact: avillas@ethz.ch

RNA sequencing variants are enriched for eQTL in cattle tissues

“…SNP precision and recall were slightly higher for WGS at 100M reads compared to the RNA-seq (Figure 2e). However, at 30M reads, the RNA-seq outperformed DNA-seq for both SNP precision and recall, although 1.6-fold DNA-seq is far below a typical variant calling depth and requires processing with low pass imputation approaches to achieve sufficiently accurate genotypes (40).…”

The efficient phasing and imputation pipeline of low‐coverage whole genome sequencing data using a high‐quality and publicly available reference panel in cattle

“…The microarray data are more prone to ascertainment bias compared with the whole genome resequencing data [2] and cannot find and evaluate new variants, which affects the accuracy of genotype imputation to a further extent. In addition, the imputation accuracy, especially in livestock, is relatively low such as correlation lower than 85% [1, 3] due to factors such as chip density and reference panel.…”