Bayesian analyses of multiple epistatic QTL models for body weight and body composition in mice

Abstract

To comprehensively investigate the genetic architecture of growth and obesity, we performed Bayesian analyses of multiple epistatic quantitative trait locus (QTL) models for body weights at five ages (12 days, 3, 6, 9 and 12 weeks) and body composition traits (weights of two fat pads and five organs) in mice produced from a cross of the F1 between M16i (selected for rapid growth rate) and CAST/Ei (wild-derived strain of small and lean mice) back to M16i. Bayesian model selection revealed a temporally regulated network of multiple QTL for body weight, involving both strong main effects and epistatic effects. No QTL had strong support for both early and late growth, although overlapping combinations of main and epistatic effects were observed at adjacent ages. Most main effects and epistatic interactions had an opposite effect on early and late growth. The contribution of epistasis was more pronounced for body weights at older ages. Body composition traits were also influenced by an interacting network of multiple QTLs. Several main and epistatic effects were shared by the body composition and body weight traits, suggesting that pleiotropy plays an important role in growth and obesity.

Figure 1

Genome-wide epistatic analysis of body weights: profiles of posterior inclusion probability, cumulative probability function, posterior means of main effect and proportion of phenotypic variance explained by main effect (PPV%). On the x-axis, outer tick marks represent chromosomes and inner tick marks represent markers. 12d, 3wk, 6wk, 9wk and 12wk represent body weights at day 12 and at 3, 6, 9 and 12 weeks of age, respectively.

Figure 2

Genome-wide epistatic analysis of fat traits under model without adjustment for body weight at 12 weeks: profiles of posterior inclusion probability, cumulative probability function, posterior means of main effect and proportion of phenotypic variance explained by main effect (PPV%). On the x-axis, outer tick marks represent chromosomes and inner tick marks represent markers. GON, SUB and FAT represent perimetrial fat pad, right hindlimb subcutaneous fat pad and the sum of the two fat pads, respectively.

Figure 3

Genome-wide epistatic analysis of fat traits under model including adjustment for body weight at 12 weeks: profiles of posterior inclusion probability, cumulative probability function, posterior means of main effect and proportion of phenotypic variance explained by main effect (PPV%). On the x-axis, outer tick marks represent chromosomes and inner tick marks represent markers. GON, SUB and FAT represent perimetrial fat pad, right hindlimb subcutaneous fat pad and the sum of the two fat pads, respectively.

Figure 4

Genome-wide epistatic analysis of organ weights under model excluding adjustment for body weight at 12 weeks: profiles of posterior inclusion probability, cumulative probability function, posterior means of main effect and proportion of phenotypic variance explained by main effect (PPV%). On the x-axis, outer tick marks represent chromosomes and inner tick marks represent markers. HRT, LIV, SPL, TES and KID represent weights of heart, liver, spleen, testis and kidney, respectively.

Figure 5

Genome-wide epistatic analysis of organ weights under model including adjustment for body weight at 12 weeks: profiles of posterior inclusion probability, cumulative probability function, posterior means of main effect and proportion of phenotypic variance explained by main effect (PPV%). On the x-axis, outer tick marks represent chromosomes and inner tick marks represent markers. HRT, LIV, SPL, TES and KID represent weights of heart, liver, spleen, testis and kidney, respectively.