QTL mapping for growth traits of pigs using random regression models

Abstract

The objective of this work was to evaluate the efficiency of random regression models (RRM) to detect quantitative trait loci (QTL) for growth traits in pigs. An outbreed F2 Piau x Commercial population was used. The efficiency of the proposed methodology for QTL detection was compared to the traditional methodology of regression interval mapping. In order to do that, RRM with random polygenic effects of permanent environment and of QTL were used with the approach of the covariance matrix identical‑by‑descent, associated to the QTL effects. The significance of QTL effects was tested by the likelihood ratio, considering the described model as being complete when there was significant QTL effect, and null, when there was not. Comparison between models was made at the positions of the markers (six microsatellite markers) and at the intermediate positions between them. The RRM allowed the detection of a significant QTL at the position 65 cM of chromosome 7 and, therefore, it was considered more efficient than traditional method, which did not identify any significant QTL in any of the evaluated phenotypes. The proposed methodology enabled the detection of QTL affecting all growth trajectory within the considered age range (from birth to 150 days).