A population-genomic approach for estimating selection on polygenic traits in heterogeneous environments

Area of Science:

• Evolutionary biology
• Population genetics
• Quantitative genetics

Background:

• Strong selection drives rapid evolution, but fluctuating selection can impede long-term evolutionary change.
• Temporal variation in selection intensity and direction impacts molecular diversity, plasticity, and ecological specialization.
• Detecting fluctuating selection is difficult due to analytical limitations.

Purpose of the Study:

• To develop and validate a novel approximate Bayesian computation (ABC) method for detecting and quantifying fluctuating selection on polygenic traits.
• To model environment-dependent phenotypic selection and its evolutionary genetic consequences.
• To provide an analytical tool for population genomic time-series data.

Main Methods:

• Developed an approximate Bayesian computation (ABC) framework to analyze population genomic time-series data.
• Proposed a model for environment-dependent phenotypic selection linked to genotype-phenotype mapping.
• Utilized simulations to assess method accuracy and precision.

Main Results:

• The ABC method provides accurate and precise estimates of fluctuating selection when the data's generative model matches the method's assumptions.
• Application to a cowpea seed beetle study showed method performance was dependent on analytical choices.
• The method successfully connects causes of variable selection to traits and genome-wide evolutionary patterns.

Conclusions:

• The proposed ABC method offers a powerful approach to study fluctuating selection, despite some limitations.
• The method aids in understanding the evolutionary genetic consequences of variable selection.
• Open-source software (fsabc) is available for implementing this analytical method.