Help Centre

The BREEDPLAN Advantage: Data Quality Checks

BREEDLAN has several data quality checks that are routinely carried out to ensure the accuracy of the data that enters the genetic evaluation. Unlike some other genetic evaluations which automatically and permanently exclude data identified by their data checks, BREEDPLAN notifies breeders when such data is flagged. This process allows breeders to check for typos, misreads and other accidental errors in the data and to correct these issues where possible.

In chronological order, the BREEDPLAN data quality checks include:

  • The breeder’s themselves. If incorrect data is submitted, the first person hurt is the breeder because their subsequent selection decisions and the reputation of their bulls may be affected. This provides strong motivation to get the recording right.
  • Automatic data submission checks include confirmation that the range of data is typical for the breed, the age of measurement is appropriate for the trait and that the animal is the correct sex for sex limited traits (e.g. male for scrotal size, female for mature cow weight).
  • All BREEDPLAN analyses conduct outlier checks.
    • An outlier occurs when a piece or pieces of performance data is beyond the range expected for a particular trait from the group of animals with which it is being compared.
    • Outliers are expected and not automatically assumed to be incorrect (despite their automatic exclusion in some other genetic evaluations).
  • The DNA going into Single-Step BREEDPLAN analyses is checked for the quality of the sample, resemblance to the reference population, the presence of duplicate profiles on the database (possible for identical twins) and checking that the pedigree and sex recorded on the breed society database matches the DNA results.
  • There are a number BREEDPLAN processes to flag data for further checks or apply fixes when potential data issues are identified.

In addition, BREEDPLAN users can be further reassured by the application of Best Linear Unbiased Prediction (BLUP) and Single-step BLUP (SSBLUP) methodologies to generate BREEDPLAN EBVs. BLUP and SSBLUP utilise many sources of information in their calculations and the effect of a single incorrect piece of information is reduced compared to genetic evaluations that do not utilise all of this information.

Click here to read more about the BREEDPLAN Advantages.