Skip to main content
ARS Home » Plains Area » Clay Center, Nebraska » U.S. Meat Animal Research Center » Genetics and Animal Breeding » Research » Publications at this Location » Publication #149934
Title: A LARGE-SAMPLE QTL STUDY IN MICE: I. GROWTH

Author
item ROCHA, JOAO - UNIV. OF NEBRASKA-LINCOLN
item EISEN, EUGENE - NORTH CAROLINA STATE UNIV
item Van Vleck, Lloyd
item POMP, DANIEL - UNIV. OF NEBRASKA-LINCOLN

Submitted to: Mammalian Genome
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 9/10/2003
Publication Date: 3/1/2004
Citation: ROCHA, J.L., EISEN, E.J., VAN VLECK, L.D., POMP, D. A LARGE-SAMPLE QTL STUDY IN MICE: I. GROWTH. MAMMALIAN GENOME. 2004. 15:83-99.

Interpretive Summary: The results of this study which had a relatively large sample size (about 1,000 F2 mice) and several growth traits, when integrated with results of previous studies, provide a general understanding of principles of QTL detection and of the genetic architecture of growth. Several primary conclusions emerge. First, numbers of QTL and magnitudes of individual effects detected are clearly dependent on sample size. Small experiments can lead to large estimates of QTL effects which are usually the result of statistical biases. Later experiments with larger sample size often considerably shrink the magnitude of estimates of effects detected in smaller experiments. Second, the distribution of QTL effects clearly does not conform with the uniform distribution which is the basis of the infinitesimal model widely assumed in animal breeding, but rather approximates an exponential model of effects (as predicted earlier by others) which nonetheless maintains an infinitesimal quality. Third, genetic control of growth derives primarily from localized genomic regions, and not from the genome as a whole. In fact, vast genomic regions seem to have little effect on growth. Other researchers have obtained similar results. Fourth, the agreement observed between most estimates of dominance effects and theoretical expectations determined by principles of classical quantitative genetics and the consistency observed for locations of many detected QTL across a wide variety of studies using different genetic backgrounds are impressive. These results support the soundness of statistical methodology that has been developed to implement QTL analyses.

Technical Abstract: Using long-term selection lines for high and low growth, a large-sample (about 1,000 F2) experiment was conducted with mice to further understand the genetic architecture of complex polygenic traits. In combination with previous work, we conclude that QTL analysis has reinforced the classic polygenic paradigms put in place prior to molecular analysis. Composite interval mapping revealed large numbers of QTL for growth traits with an exponential distribution of magnitudes of effects and validated theoretical expectations regarding gene action. Of particular significance, large effects were detected on Chromosome 2. Regions on Chromosomes 1, 3, 6, 10, 11 and 17 also had loci with significant contributions to phenotypic variation for growth. Despite the large sample size, average confidence intervals of about 20 cM exhibit the poor resolution for initial estimates of QTL location. Analysis with genome wide and chromosomal polygenic models revealed that, under certain assumptions, large fractions of the genome may contribute little to phenotypic variation for growth. Other primary observations from this study were: few epistatic interactions among detected QTL, little statistical support for gender-specific QTL, and significant age effects on genetic architecture.