Selection of genetically resistant animals is one alternative to reduce the negative impact of gastrointestinal nematodes (GIN) on sheep production. The aim of this study was to identify genomic regions associated with GIN resistance in Corriedale sheep by single-step genome-wide association studies (ssGWAS) using 170, 507 and 50K single nucleotide polymorphisms (SNPs). Analysis included 19,547 lambs with faecal egg counts (FEC) records, a pedigree file of 40,056 animals and 454, 711 and 383 genotypes from 170, 507 and 50K SNPs, respectively. Genomic estimated breeding values (GEBV) were obtained with single-step genomic BLUP methodology (ssGBLUP), using a univariate animal model, which included contemporary group, type of birth and age of dam as class fixed effects and age at FEC recording as covariate. The SNP effects as wells as p-values were estimated with POSTGSF90 program. Significance level was defined by a chromosome-wise False Discovery Rate of 5%. Significant genomic regions were identified in chromosomes 1, 3, 12 and 19 with the 170 SNP set, in chromosomes 7, 12 and 24 using the 507 SNP chip and only in chromosome 7 with the 50K SNP chip. Candidate genes located in these regions, using Oar_v4.0 as reference genome, were TIMP3, TLR5, LEPR and TLR9 (170 SNPs), SYNDIG1L and MGRN1 (507 SNP chip) and INO80, TLN2, TSHR and EIF2AK4 (50K SNP chip). These results validate genomic regions associated with FEC previously identified in Corriedale and other breeds and report new candidate regions for further investigation.

One control strategy for gastrointestinal nematodes (GIN) is genetic selection. This study´s objective was to compare eggs per gram of feces (FEC) and fiber diameter (FD) estimated breeding values (EBV) and genomic EBV (GEBV) in Corriedale breed. Analysis included 19547 lambs with data, and 454, 711 and 383 genotypes from 170, 507 and 50K SNP chips, respectively. A univariate animal model was used for EBV and GEBV estimation, which included contemporary group, type of birth and dam age as fixed effects, and age at recording as covariate. Differential weights (α) were considered in the genomic relationship matrix (G), and the best fit models were identified using Akaike´s Information Criterion (AIC), which were later used for GEBV and accuracies estimation. The use of α only impacted on low density SNP chips. No differences were observed in mean accuracies for the whole population. However, in the genotyped subgroup accuracies increased by 2% with the 170 SNP chip (α=0.25), and by 5% (α=0.5) and 14% (α=0.75) with the 507 SNP chip. No differences were observed in FD EBV and GEBV mean accuracies. These results show that it is possible to increase GEBV accuracies despite the use of low-density chips.