R.F. Veerkamp and Y. de Haas (eds) Proceedings of 12 th World Congress on Genetics Applied to Livestock Production (WCGALP) 2988 DOI: 10.3920/978-90-8686-940-4_725, © A. Kasap et al. 2022 725. Estimation of linkage disequilibrium in Pag and Istrian sheep breed: towards genomic optimum contribution selection A. Kasap 1* , J. Ramljak 1 and M. Špehar 2 1 University of Zagreb Faculty of Agriculture, Svetošimunska 25, 10000 Zagreb, Croatia; 2 Croatian Agency for Agriculture and Food, Svetošimunska 25, 10000 Zagreb, Croatia; akasap@agr.hr Abstract Knowledge of the genetic architecture of the population is important to predict the potential of genomic selection. e strength of association between genome wide single nucleotide polymorphisms (SNPs) and unknown quantitative trait loci (QTL), measured by linkage disequilibrium (LD), represents the corner stone of genomic selection. e aim of this study was to estimate LD in two Croatian native dual purpose (milk-meat) sheep breeds (Pag sheep, n=719 and Istrian sheep, n=749) using the Illumina Ovine 50K SNP chip. e main goal in these populations is to improve dairy yield traits, preferably via genomic optimum contribution selection. e estimated LD was low (r 2 Pag =0.04, D’ Pag =0.28, r 2 Istrian =0.06, D’ Istrian =0.39,) with very steep LD decay in both breeds. e results obtained are promising for conservation of these breeds, but less promising for the success of genomic selection. In order to maximize benefits of genomic information, single step genomic BLUP should be applied. Introduction Istrian and Pag sheep are Croatian native dairy orientated breeds with great economical, traditional and cultural importance. In order to preserve them ‘in situ’, it is necessary to permanently increase their productivity. Small breeds under selection are under risk of genetic erosion, but low productivity leads them towards extinction due to non-competitiveness with highly selected breeds (FAO, 1998). Balancing between selection gain and loss of genetic variability is a challenging but feasible task, best achieved by following the principles of optimum contribution selection (OCS) (Wellmann et al., 2021; Wooliams et al., 2015). High-density SNP arrays have been used as an important source of information for selection (Van Raden, 2020) and conservation (Windig and Engelsma, 2010) in many livestock populations. However, benefits of genomic information in selection vary between populations, and depend heavily on population genetic parameters such as linkage disequilibrium (LD). Association between marker alleles (SNP) and causative alleles (QTL) directly affects the accuracy of genomic selection (Meuwissen et al., 2001) and estimation of genetic diversity parameters (Boettcher et al., 2010, Li et al., 2011). Wishing to upgrade the existing selection strategy to genomic optimum contribution selection, we aimed to estimate the genome- wide levels of LD. e obtained results will serve as a base to ascertain the benefits of potentially novel selection strategy in these breeds. Materials & methods Samples for genotyping were collected with the permission of the Bioethical committee for protection and welfare of animals at University of Zagreb Faculty of Agriculture. Ear tissue from 719 Istrian sheep (13 flocks) and 749 Pag sheep (11 flocks) were taken with Allflex tissue sample unit applicator. Animals were originally genotyped for 52,152 SNPs, using the Illumina OvineSNP50K BeadChip. Only autosomal SNPs were included in the analysis aſter quality control parameters for missingness per animal, missingness per SNP, and minor allele frequency set to 0.9, 0.95, and 0.01, respectively. SNPs were also removed if they deviated from the Hardy-Weinberg equilibrium (P-value cut-off set to 1×10 -6 ). A total of 693 Istrian (44,743 SNPs) and 713 Pag sheep (46,685 SNPs) passed the above control. e linkage disequilibrium (LD) between pairs of SNPs was calculated with the PLINK (version 1.9) (Purcell et al., 2007; Chang et al., 2015) using the two most commonly used statistics r 2 and D’. Post hoc statistical analysis of genome wide LD was https://www.wageningenacademic.com/doi/pdf/10.3920/978-90-8686-940-4_725 - Saturday, February 11, 2023 12:05:30 PM - IP Address:35.170.81.163