Sequencing & Pangenomes

Viability of Low-Coverage Nanopore Sequencing for Genomic Characterisation and Imputation Accuracy in an Endangered Pig Breed

---

Abstract

BackgroundGenotyping arrays are the default technology for routine genomic characterisation in breeding and conservation. They are highly efficient in terms of precision, speed and cost, but also afflicted with ascertainment bias and limited to a preselected SNP set. Cost efficient use often requires a minimum number of samples, which makes them unattractive for scenarios where small breeds or just small batches must be characterised. Long-read sequencing promises to overcome some of these issues, as it can be done even under field conditions and for very low numbers of samples. Additionally, it holds the potential for enhanced detection of structural variation and methylation. Unfortunately, it is afflicted with substantial cost and computational burdens. Therefore, promising strategies are based on low-coverage sequencing of multiplex libraries from multiple animals in one run. As the individual coverage is insufficient for reliable variant calling, imputation methods, either with or without reference set must be applied to numerous animal sets. ObjectiveWe aim to evaluate, if low-coverage long-read sequencing with following imputation can serve as a replacement for array genotyping in a local German pig breed. We are especially interested, which individual coverage would be sufficient for reliable genotyping. Material & MethodsWe use barcoded Nanopore sequencing to generate low coverage sequences of an endangered German pig breed, the Swabian-Hall pig. Six animals were sequenced per flowcell, using the Native barcoding protocol of Oxford Nanopore (ONT). Currently 185 animals are finished. Samples are processed with a Dorado/ Minimap based pipeline. Stitch is used as a reference free imputation tool, and Quilt as a reference-based alternative, to generate genotypes for the positions of a commercially available 80 k Illumina array. The reference set consists of a variant set called from publicly available short-read data of ~80 individual pigs representing global diversity. All samples were also genotyped with the respective 60k array. ResultsGenerated depth of coverage per sample ranged from 0.1 to 15 X. The allele frequency spectrum of the Stitch call-set resembles that of the array, while the Quilt set shows an excess of monomorphic lost and fixed variants compared to the array. Genotype concordance is generally higher for Stitch ( > 95 %, apart from few outlier positions) than for Quilt (majority > 80 %), distinctly at average allele frequencies. Stitch maintains concordance > 95 % for coverages as low as 1 X, and still has 90 % concordance for lowest coverages, while the Quilt approach showed decrease in concordance from 5 X downwards. ConclusionLow-coverage long-read sequencing with ONT can be an alternative to array genotyping, even in small breeds where use of commercial arrays might be restricted, and offers a fast independent solution for genomic characterization.

Keywords: 2026

How to Cite:

geibel, J., Hinrichs, D., Kistner, L., Petig, M., Reimer, C., Vollmer, L., Weigend, A., Weigend, S., Weishaar, R. & Worat, C., (2026) “Viability of Low-Coverage Nanopore Sequencing for Genomic Characterisation and Imputation Accuracy in an Endangered Pig Breed”, World Congress on Genetics Applied to Livestock Production Digital Archive 2026(1): 2286932. doi: https://doi.org/10.31274/wcgalp.24172

Rights: 1