Implications of cryptic correlations caused by interacting traits for quantitative genetics

Authors: Bennett, Gary

Submitted to: International Conference on Quantitative Genetics
Publication Type: Abstract Only
Publication Acceptance Date: 10/5/2019
Publication Date: N/A
Citation: N/A

Interpretive Summary:

Technical Abstract: Genetic correlations ascribed to pleiotropy or linkage are expected to change slowly due to linkage breakdown or changing allele frequencies. Correlations are expected to accurately predict short term correlated responses to selection. Primary traits may interact to produce derivative traits as in some numerical models of traits. Derivative traits may be more important and easily measured than primary traits or the primary traits and their interaction may be unknown or unmeasurable. Resulting correlations can be considered cryptic. An example of a modeled interaction is when a derivative trait is the minimum of primary traits. Several reproduction traits are modeled this way and body composition is often modeled as an interaction of feed intake and innate growth potential. A derivative trait (minimum of two uncorrelated primary traits) was simulated. Two restricted selection indexes using one primary trait and the derivative trait were calculated to change one trait and not the other while the second primary trait was ignored (considered cryptic). Four 1-generation selections were simulated to either increase or decrease each of the two restricted indexes. Genetic correlation changed from 0.66 to 0.46, 0.65, 0.80, and 0.49. Only one of four restricted changes was close to zero. Predicted and actual changes in correlated traits and in correlations following natural or artificial selection need to be carefully interpreted. In artificial selection, strategies that use experimental and industry populations but do not depend on estimated correlations are needed to validate predicted correlated changes. USDA is an equal opportunity employer and provider.