Results and Discussion

MAF variation
The MAF for each SNP was calculated from the genotypic data and is presented in Table 2. Analysis of 7,045 markers revealed that the mean MAF was 0.23, 0.22, 0.21, 0.21, 0.23, and 0.29 for Ambo, Arsi, Borana, Danakil, Horro, and Hanwoo cattle, respectively. A significant difference was observed between the Ethiopian indigenous cattle populations and Hanwoo (p < 0.001). The average value for Hanwoo was 32% higher than the value for Ethiopian cattle populations. The overall MAF in the present finding for indigenous Ethiopian cattle breeds is higher than an earlier report for indicine breeds but lower than the average value reported for most taurine breeds [3]. The value found for Hanwoo cattle was comparable with the previous results for taurine breeds but higher than the reported average for the same breed [10, 12]. This could be attributed to the different marker density used, and a higher density marker (Illumina Bovine 50K SNP BeadChip) was used in the previous study. The higher values for taurine breeds can be explained by the fact that SNP loci used in this study were detected in taurine breeds, and their average minor allele frequency was much lower in indicine breeds.
Across the Ethiopian cattle populations, common variant MAFs (≥0.10 and ≤0.5) accounted for 74% of the total SNPs (Table 3). Hanwoo cattle displayed a higher proportion of common variant SNPs (90%). Rare variant MAFs were observed in about 9% and 4% of markers in the Ethiopian and Hanwoo breeds, respectively. The relatively higher proportion of fixed alleles in Ethiopia cattle populations could be attributed to the inbreeding as a result of uncontrolled breeding practices. A common variant MAF proportion of 83% was reported in sheep breeds [5]. The distribution of MAFs revealed that nearly 34.08% of the markers in Ethiopian cattle populations and 53% in Hanwoo cattle displayed a high degree of polymorphisms (0.3 ≤ MAF ≤ 0.5). Yet, this value was higher than an earlier report both for Hanwoo (25%) and Holstein (28%) [12]. Much higher MAFs were detected in Bos taurus breeds than in Bos indicus breeds [3]. The proportion of fixed markers varied among the sampled populations. Among the Ethiopian cattle populations, the highest (8.96%) and lowest (6.30%) were obtained in the Borana and Ambo populations, respectively. The Hanwoo breed displayed a lower level of fixed SNPs (1.10%).
On average, 83.36% of the markers displayed polymorphisms (MAF > 0.05) within Ethiopian cattle populations, and the remaining 16.64% of the SNPs considered as monomorphic. The Hanwoo breed showed the highest proportion of polymorphisms (94%) (Fig. 3). The level of polymorphic SNPs in the present study was higher than what was previously reported for taurine breeds [10]. But, it was in close agreement with the value reported for sheep breeds using genome-wide SNPs [5]. Similarly, Gautier et al. [13] reported a lower rate of polymorphisms using 696 SNPs localized in BTA03 from European and West African cattle breeds (94.6% and 81.8%), respectively. The higher observed polymorphisms in Hanwoo cattle could explained by the fact that most of the bovine sequence data available in the database are from Bos taurus breeds.

Genetic diversity among Ethiopian cattle populations
An examination of the variability within breeds was carried out to compare the level of hererozygosity between the populations, and the results are presented in Table 4. Across the sampled Ethiopian cattle populations, the mean observed and expected heterozygosities were 0.314 ± 0.001 and 0.313 ± 0.001, respectively. The mean values were lower than the values reported from a microsatellite markers analysis [1]. However, they were in concurrence with the values reported using SNPs for the Angus breed (0.332) [14] and for sheep breeds [5].
The overall estimated F-statistics were FIS = -0.003 (within-population inbreeding estimates) and FIT = 0.016 (total inbreeding). Among the Ethiopian cattle populations, genetic differentiation (FST) was estimated at 0.019. Genetic characterization of Ethiopian cattle breeds using microsatellites reported within-population inbreeding of 0.071, total inbreeding of 0.083, and genetic differentiation of 0.013 [1]. The FST for Ethiopian cattle populations was low, which could be due to their common ancestral origin.
There were loci significantly deviating from HWE. Overall, significant deviation from Hardy-Weinberg proportions (p < 0.05) was observed for 405 markers in Ethiopian cattle populations (Ambo, 441; Arsi, 475; Borana, 314; Danakil, 314; and Horro, 481) and in 549 of the markers in Hanwoo. The highest number of SNPs that departed from HWE was observed in Hanwoo cattle, whereas within the Ethiopia cattle populations, Borana displayed lower proportions of markers (SNPs) deviating from HWE. The variation in proportion of SNPs displaying deviation from HWE among the sampled populations could be explained by selection pressure and population structure (stratification, admixture).
In conclusion, a significant difference was detected for the rate of polymorphisms and MAFs between the Hanwoo and the Ethiopian cattle populations. The level of variation identified in this particular study highlights that these markers can be potentially used for genetic studies in African cattle populations. The high within-breed diversity observed in Ethiopian indigenous cattle populations could be an opportunity and allow improvements through within-population selection schemes.