Integrating Pool-seq uncertainties into demographic inference.

Area of Science:

• Population genetics
• Evolutionary biology
• Bioinformatics

Background:

• Next-generation sequencing of pooled samples (Pool-seq) is widely used for population diversity analysis.
• Pool-seq data can be noisy due to factors like unequal individual contributions, limiting its use in evolutionary history reconstruction.
• Existing methods often do not adequately address Pool-seq specific error sources.

Purpose of the Study:

• To develop a novel Approximate Bayesian Computation (ABC) method for inferring demographic history from Pool-seq data.
• To explicitly model and account for sources of noise inherent in Pool-seq data.
• To assess the utility of Pool-seq data for distinguishing between different scenarios of ecotype formation and inferring demographic parameters.

Main Methods:

• Developed a novel Approximate Bayesian Computation (ABC) method incorporating Pool-seq error models.
• Jointly modeled Pool-seq data, demographic history, and selection effects.
• Employed computationally efficient simulations using subsets of loci and relative statistics/parameters.

Main Results:

• The ABC method successfully infers demographic history parameters from Pool-seq data, accounting for technical errors.
• Simulation studies confirmed Pool-seq data's ability to differentiate ecotype formation scenarios (single vs. parallel origin).
• Application to *Littorina saxatilis* data indicated ecotype divergence predated local colonization, persisting despite gene flow.

Conclusions:

• Demographic modeling and inference are feasible and reliable using Pool-seq data with the proposed ABC method.
• The approach provides a valuable tool for understanding the genetic basis of adaptation in natural populations.
• This work advances the development of null models for analyzing population genomics data, particularly from pooled samples.