Identification of key contributors in complex population structures.
Authors: Neuditschko Markus, Raadsma Herman W, Khatkar Mehar S, Jonas Elisabeth, Steinig Eike J, Flury Christine, Signer-Hasler Heidi, Frischknecht Mirjam, von Niederhäusern Ruedi, Leeb Tosso, Rieder Stefan
Journal: PloS one
Summary
# Editorial Summary Understanding which animals truly shape a population's genetic structure is fundamental for making sound breeding decisions, managing conservation programmes, and selecting animals for costly genome sequencing projects. Neuditschko and colleagues developed an innovative three-step method based on eigenvalue decomposition of genomic relationship matrices to identify key genetic contributors and visualise complex population substructures; they validated this unsupervised approach across four datasets including horse and cattle populations of varying complexity (1,077 and 2,457 individuals respectively). Compared to established selection strategies for building reference populations for genotype imputation—such as marginal gene contribution calculations or genetic relatedness optimisation—their key contributor method achieved the highest phasing accuracies, suggesting superior utility for downstream genomic work. The approach proved particularly valuable in identifying fine-scale population substructures even when no obvious influential ancestors were present, revealing the hidden genetic architecture in complex, admixed breeding populations that conventional methods might overlook. For practitioners engaged in selective breeding, conservation planning, or genomic research, this integrated approach offers a more sophisticated tool than currently standard methods for both characterising population complexity and making strategic decisions about which animals warrant investment in genome sequencing or serve best as reference animals for imputation.
Read the full abstract on PubMed
Practical Takeaways
- •This population genetics tool helps breeders identify which individuals contribute most to genetic diversity, improving selective breeding decisions and conservation strategies
- •For operations planning genomic testing or imputation projects, selecting key contributors as reference individuals yields better accuracy than traditional methods
- •The integrated approach provides clearer visualization of population structure and admixture, supporting informed management of complex pedigrees in horse and cattle operations
Key Findings
- •A novel three-step approach using Eigenvalue Decomposition identified key genetic contributors and high-resolution population substructures in horse (1,077 individuals) and cattle (2,457 individuals) populations
- •Selection of key contributors provided higher phasing accuracy for genotype imputation compared to marginal gene contribution and genetic relatedness optimization methods
- •The unsupervised method successfully identified all known key contributors in simulated and experimental sheep populations, validating its effectiveness across disparate datasets