SUMMARY OF ARTICLE

ABSTRACT

Genomic selection has bought a remarkable impact in dairy cattle improvement programs, particularly in navigating the complexities and costs associated with longer generation intervals and sex-limited traits pertaining to conventional pedigree-based selection schemes. The adoption of genomic selection in developed countries like United States, Australia, Netherland, New Zealand and Canada has doubled the rate of genetic progress within the dairy industry. Meanwhile, the implementation of genomic selection in developing countries including India is still at a budding stage due to various constraints. The main hinderance for implementation of genomic selection in India includes limited pedigree, challenges involved in construction of a sizable reference population in numerically small breeds, and the small animal holding system of India, which leads to the effect of uneven management and uncontrolled mating. On the other hand, the fact remains that our country has one of the highest populations of cattle and number of breeds available within the species are also very large when compared to European and American continents. Under such conditions, selection programme based on the genomic information of animals would be useful for early selection of animals with increased accuracy. Nevertheless, it is important to recognise that addressing the limitations is a prerequisite for successful adoption of genomic selection. Specifically, at NDRI, pilot studies were conducted to address the challenge of limited pedigree and poor availability of resources using simulated data for Indian conditions. In order to cut down cost on genotyping, different selective genotyping strategies were evaluated and the impact of selective genotyping for creation of reference population was also studied which revealed that employing a twotailed selective genotyping approach with ssGBLUP resulted in higher prediction accuracy with reduced bias. Furthermore, different breeding value prediction models along with selective genotyping strategies were compared to develop a genomic prediction strategy (pipeline) capable of accurately estimating unbiased GEBV for sex-limited traits with shallow pedigree information (i.e, limited pedigree). The findings highlighted that use of ssGBLUP with two tailed selectively genotyped all female reference in shallow pedigree scenarios, resulted in unbiased and accurate GEBV for sex-limited traits, when resources are limited.