Cluster analysis and breed prediction using genomic data in a multibreed beef cattle population

• Grant number: 18/14035-4
• Support Opportunities: Scholarships in Brazil - Scientific Initiation
• Start date: October 01, 2018
• End date: September 30, 2019
• Field of knowledge: Agronomical Sciences - Animal Husbandry - Genetics and Improvement of Domestic Animals
• Principal Investigator: Ricardo Vieira Ventura
• Grantee: Lucas Tassoni Andrietta
• Host Institution: Faculdade de Medicina Veterinária e Zootecnia (FMVZ). Universidade de São Paulo (USP). São Paulo , SP, Brazil
• Associated research grant: 16/19514-2 - Development of a genomic database for Nellore cattle and computational tools for implementing large-scale studies, AP.JP

Abstract

Predicting the breed composition of animals derived from a multi breed beef cattle population can help to better understand how the current population have developed, as well as to identify the most suitable prediction method to obtain breeding values for the suggested set of animals. Determining breed composition of animals that are thought to be purebred may give us immediate response for discrepant genetic evaluation results obtained for certain animals. Currently, several research groups have conducted genome projects in cattle and data sharing became a reality at research level, even across countries. One of the main challenges regarding the use of shared datasets from different project sources is related to the conversion and standardization of the genetic markers due to the different file formats, number of the markers in panels and adopted platform technologies. The main objective of our investigation is to evaluate clustering approaches and methods to predict breed composition using genomic data shared from Canadian partners (AgSights, previously named Beef Improvement Opportunities) and also from a dataset obtained from a public source in the web. Before carrying out the referred studies, this study will also investigate different strategies for combining the datasets, as well as will quantify the problems for each planned solution.