Female Reproduction Phenobank and Validation Herds Final Report

Summary

The female reproduction PhenoBank project was undertaken to collect datasets that are informative for building a reference population, which will facilitate genomic selection in Northern Australian cattle. The datasets in this project were collected with support from beef producers that volunteered the performance records of their cows, together with a hair or tissue sample for DNA analyses. The used chip technology for genotyping DNA, followed by data analyses discovered associated DNA markers and computed the accuracy of genomic selection for reproduction traits. The traits used were relatively easy to measure, to facilitate future adoption. They were based on pregnancy outcomes of the first and second breeding seasons. Heifer pregnancy outcomes and their rebreeding capacity were heritable traits, within the expectations for female reproduction traits. Prediction accuracies for these traits demonstrates their potential in terms of using DNA data to improve reproductive efficiency.

Objectives

Objective 1: Create the Female Reproduction PhenoBank. Multiple datasets were integrated into one database with a standardised vocabulary to define female reproduction traits. Custom software was developed to integrate the multiple datasets of phenotypes and genotypes and provide easy, yet controlled access for end-users (i.e., PhenoBank collaborators) of the data. PhenoBank currently has integrated and matched female reproduction phenotypes and genotypes for 10,507 cows. This level of genome-wide data for tropical breeds is a world-class resource.

Objective 2: Identification of functional mutations for female reproduction traits. The PhenoBank database was used to identify mutations associated with female reproduction traits, and confirmed that female reproduction phenotypes are polygenic traits, with multiple associated mutations, all with relatively small effects individually.

Objective 3: Validation and Industry Herds. The mutations identified in Objective 2 were used to inform the subsequent genomic selection analyses and in this context, they were validated. This analysis adds evidence, or proof of concept, to the idea of using associated mutations to enhance the predictions in industry datasets. The measurable output of this work is estimated accuracies of genomic Estimated Breeding Values (gEBVs) across independent cattle populations.

Objective 4: Database protocol. The establishment of an appropriate infrastructure for PhenoBank data storage required standardised protocols for importing, manipulating, and exporting data, including meta-data to ensure source identification and appropriate use. A PhenoBank manual was developed.

Key findings

- Both windows and web-based front-end applications were developed for Phenobank to access Livestock Information Platform (LIP). The applications were developed in consultation with all PhenoBank researchers, and the software improved in every interaction.
- Genomic models for the two traits, PREG1 (pregnancy outcome of the first and the second breeding seasons), estimated heritabilities of 0.17 and 0.21 respectively. These heritabilities confirm the possibility of targeting these industry-based traits for selection.
- The accuracies of genomic selection improved when GWAS results were used to inform the predictions.

Benefits to industry

The northern beef industry now has a solid platform to underpin substantial genetic gains in female reproduction traits. The database created and the data assembled can be made available to cattle geneticists in Australia to improve their predictions of genetic merit for female reproduction traits. PhenoBank resources can inform the use of genomic data for breeding tropically adapted cattle. The main audiences for this research are beef producers, especially breeding herds of tropically adapted cattle in Northern Australia, and geneticists seeking to provide information and service to these producers.

MLA action

It is recommended
1) to continue research in this area by using the full PhenoBank data
2) to further develop the PhenoBank database so that it can store sequence level data (i.e. millions of DNA markers per animal)
3) to investigate if PhenoBank resources can be integrated with other projects to further enhance the reference population available for female reproduction traits
4) to provide the herd specific phenotypes and genotypes back to the producer who volunteered their data and samples

Future research

Software upgrades that allow storage and usage of sequence level data will enhance discovery activities. There is also potential to explore genetic correlations between the traits in Phenobank and those measured in other projects, such as Repronomics and Smart futures. Potentially, these projects could become another source of data for Phenobank to ultimately generate more robust predictions. Likewise, PhenoBank data could be made available to producers, researchers and genetic improvement programs so that data collected with MLA funds benefits everyone in the industry.