Skip to main content

Table 1 Composition of three different mixed breed training (reference) sets, and several validation sets chosen to represent different levels of relatedness to the training sets

From: Exploiting biological priors and sequence variants enhances QTL discovery and genomic prediction of complex traits

Training set: description

Training set: total

Training set: number per breed

Validation sets: in order of decreasing relatedness to the Training set

“DANZ” bulls of Dutch, Aust & N. Zealand origin with real genotypes and real phenotypesa

8920

7371 Holstein

1438 Jersey

111 Aust. Red

1. 869 Red Holstein bulls

2. 655 Australian Red cows

“AUS” Australian bulls & cows with real genotypes and real phenotypesa

16,214

11,527 Holstein:

3049 bulls, 8478 cows.

4687 Jersey:

770 bulls, 3917 cows.

1. 869 Red Holstein bulls

2. 655 Aust. Red cows

“AUS-Sim” Subset of above AUS set, with real genotypes and simulated phenotypes

10,314

7991 Holstein

2323 Jersey

1. 262 Holstein bulls only

2. 3940 Holstein bulls & cows

3. 869 Red Holstein bulls

4. 885 Aust. Red bulls & cows

  1. aphenotypes were milk, protein and fat yield: in the case of bulls these are daughter averages from progeny test and all phenotypes were corrected for known fixed effects