Sequencing & Pangenomes

Leveraging low-depth whole genome sequencing and imputation to identify semen cross-contamination

---

Abstract

Ensuring sire purity in bovine semen samples is critical for breeding programs and genomic selection. Cross-contamination - whether through mislabeling or inadvertent mixing of semen from different sires - poses significant risks to genetic progress and economic outcomes. Conventional quality control methods, such as array genotyping, often lack the sensitivity to detect low-level cross-contamination, necessitating more robust genomic approaches. We developed a sensitive methodology using low-depth whole genome sequencing (LD-WGS) combined with imputation for cross-contamination detection in commercial semen samples.Variants from LD-WGS data (~0.5à— coverage) were imputed to high-density genotypes using a large haplotype reference panel. For each sample, a deviation metric was calculated against the expected baseline genotype of each animal, enabling quantitative evaluation of sample purity and detection of non-target genetic material. Modeled cross-contamination scenarios were generated by mixing reads from non-target sires at proportions ranging from 1% to 20%, and detection thresholds were established based on deviation distributions.Results indicate that LD-WGS coupled with imputation reliably identifies cross-contamination levels as low as 1%, whereas array-genotype-based concordance metrics applied to the same modeled contamination scenarios did not consistently detect contamination below approximately 10%. Beyond cross-contamination detection, LD-WGS was also applied to validate gender purity in sex-sorted samples, using the depth ratios of chromosome-specific reads to confirm that the proportion of X- or Y-bearing sperm matched the intended sorting specification. This genomic check provides an independent verification of sorting machine performance, ensuring that sexed semen meets expected purity standards for breeding programs.We also explored the cumulative use of LD-WGS reads across multiple runs to achieve higher effective coverage for each individual, highlighting its potential for research applications such as comprehensive variant calling across the entire genome, enabling detection of novel mutations to enhance genomic selection, supporting GWAS and QTL analyses, and fine mapping of causal variants. These findings underscore LD-WGS as a cost-effective and scalable solution for ensuring contamination-free samples and validating sex-sorting accuracy, while providing powerful tools for genomic research. However, successful implementation is not straightforward - it requires a robust reference panel and well-defined baseline metrics for each sire to achieve reliable detection and minimize false positives.

Keywords: 2026

How to Cite:

Deeb, N., Silva, V., Su, H., Cedraz, H., Hanson, S., Arendt, D., Mileham, A. & Ross, P., (2026) “Leveraging low-depth whole genome sequencing and imputation to identify semen cross-contamination”, World Congress on Genetics Applied to Livestock Production Digital Archive 2026(1): 2285921. doi: https://doi.org/10.31274/wcgalp.23821

Rights: 1