Inferring the potentially complex genetic architectures of adaptation, sexual dimorphism and genotype by environment interactions by partitioning of mean phenotypes

Abstract

Genetic architecture fundamentally affects the way that traits evolve. However, the mapping of genotype to phenotype includes complex interactions with the environment or even the sex of an organism that can modulate the expressed phenotype.Line-cross analysis is a powerful quantitative genetics method to infer genetic architecture by analysing the mean phenotype value of two diverged strains and a series of subsequent crosses and backcrosses. However, it has been difficult to account for complex interactions with the environment or sex within this framework. We have developed extensions to line-cross analysis that allow for gene by environment and gene by sex interactions. Using extensive simulation studies and reanalysis of empirical data, we show that our approach can account for both unintended environmental variation when crosses cannot be reared in a common garden and can be used to test for the presence of gene by environment or gene by sex interactions. In analyses that fail to account for environmental variation between crosses, we find that line-cross analysis has low power and high false-positive rates. However, we illustrate that accounting for environmental variation allows for the inference of adaptive divergence, and that accounting for sex differences in phenotypes allows practitioners to infer the genetic architecture of sexual dimorphism. K E Y W O R D Senvironmental effect, epistasis, G × E, line-cross analysis, sexual dimorphism

The March of the Beetles: Epistatic Components Dominate Divergence in Dispersal Tendency in Tribolium castaneum

“…We suggest that these findings show that it is unlikely that the large role of epistasis that we infer is simply from dispersion among our parental lines. Furthermore, the relatively small role of epistasis that we infer from dispersion points to an added strength of accounting for model selection uncertainty during LCA not previously documented ( Blackmon and Demuth 2016 ; Armstrong et al 2019 ).…”

“…Models that exhibit this characteristic are dropped from the analysis and parameters are estimated by the remaining models. Previous simulation studies indicate that this does not lead to significant bias or loss of power in the inference of composite genetic effects under a model averaging approach ( Blackmon and Demuth 2016 ; Armstrong et al 2019 ). The AIC score for each evaluated model was recorded and we constructed a 95% confidence set of models that were used to produce model-averaged results that account for model selection uncertainty ( Burnham and Anderson 2002 ).…”

“…We used the software SAGA 2.0 to evaluate all possible models for the genetic architecture of our trait ( Armstrong et al 2019 ). Because we have a total of 8 lines, we limited the space of possible models to those with 7 or fewer parameters (composite genetic effects).…”

Wright was right: leveraging old data and new methods to illustrate the critical role of epistasis in genetics and evolution

“…Models that exhibit this characteristic are dropped from the analysis, and parameters are estimated by the remaining models. Previous simulation studies indicate that this does not lead to significant bias or loss of power in the inference of composite genetic effects under a model-averaging approach ( Armstrong et al, 2019 ; Blackmon & Demuth, 2016 ). The AICc (small sample size corrected version of AIC) score for each evaluated model was recorded, and we constructed a 95% confidence set of models that were used to produce model-averaged results that account for model selection uncertainty ( Burnham & Anderson, 2002 ).…”

“…We used the software SAGA 2.0 to evaluate all possible models for the genetic architecture of our traits with the exceptions described below that reduce the size of model space for a handful of datasets ( Armstrong et al, 2019 ). In the analyses of each dataset, we limited the space of possible models to be fit to those with one fewer parameter (composite genetic effects) than the number of cohorts included in the experiment, but we set a maximum number of parameters to seven.…”