Inferring domestic goat demographic history through ancient genome imputation

Authors: ERVEN, JOLIJN - University College Dublin; ETOURNEAU, ALICE - Inrae; MASHKOUR, MARJAN - National Center For Scientific Research Demokritos; Neupane, Mahesh; BARDOU, PHILLIPE - Collaborator; STELLA, , ALESSANDRA - Collaborator; TALENTI, ANDREA - Roslin Institute; MASIGA, CLET - Tropical Institute Of Development Innovation (TRIDI); Van Tassell, Curtis; CLARK, EMILY - Roslin Institute; POMPANON, FRANCOIS - Universite Grenoble Alpes; COLLI, LICIA - Università Cattolica Del Sacro Cuore; AMILLIS, MARCEL - Autonomous University Of Barcelona; MILANESI, MARCO - University Of Tuscia; CREPALDI, PAOLA - University Of Milano; CONSORTIUM, THE VARGOATS - Collaborator; SERVIN, BERTRAND - Inrae; Rosen, Benjamin; TOSSER-KLOPP, GWENOLA - Inrae; DALY, KEVIN - University College Dublin

Submitted to: Genome Biology and Evolution
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 9/18/2025
Publication Date: N/A
Citation: N/A

Interpretive Summary: Goats were among the earliest tended animals, making them ideal to explore results of domestication. Understanding how selection affects animals helps to anticipate how future changes in livestock will change them in the future. In this study, we used animals from current global goat populations to compare with ancient relatives of goats to determine impacts on genetic diversity and inbreeding. However, a challenge in analysis of ancient genetic data is the relatively low DNA coverage available for most ancient samples. To understand that limitation in ancient data, we used high quality goat genomes randomly thinned from 1,372 individuals. Results suggested a large effect of the initial dispersal of managed herds. These findings provide insights into the effect of early herd management on population structure and confirm the utility in leveraging low coverage ancient genomes. Studies like these are crucial for predicting the future health and resilience of livestock.

Technical Abstract: Goats were among the earliest managed animals, making them an ideal model to explore evolutionary consequences of domestication. However, a challenge in ancient genetic analysis is the relatively low genome coverage for most samples, limiting analysis to pseudohaploid genotypes. Genotype imputation offers potential to alleviate this limitation by improving information content and accuracy in low coverage genomes. To test this we used published high coverage (>8X) goat palaeogenomes, imputing downsampled genomes using the VarGoats dataset (1,372 individuals) as a reference panel. Measuring concordance between imputed and high coverage genotypes, we find high concordance after filtering for common (>5%), high confidence variants, with 0.5X genomes reaching >0.97 concordance. There is a trade-off between coverage, genotype probability (GP) thresholds, and genotype recovery, where higher coverage and more lenient GP thresholds result in higher recovery, and a reduction in heterozygous false positive rates with stricter thresholds. We then imputed 36 goat palaeogenomes with >=0.5X coverage to examine runs-of-homozygosity (ROH) and identity-by-descent (IBD) patterns. Using a novel approach combining ROH profiles across tools, we find that among Neolithic goats, ROH increases with distance from the Zagros Mountains, suggesting a large effect of the initial dispersal of managed herds. Inbreeding levels decrease across Southwest Asia in later periods. IBD mirrored this pattern, with less relatedness in the early herding site of Ganj Dareh compared to higher relatedness in goats from later in the dispersal process. These findings provide insights into the effect of early goat management on demography, and confirm the utility of imputation in leveraging low coverage palaeogenomes.