Demogenomic inference from spatially and temporally heterogeneous samples

Area of Science:

• Population Genetics
• Bioinformatics
• Computational Biology

Background:

• Genomic samples are often collected across different locations and time periods, leading to population structure.
• Ignoring this heterogeneity in genomic data can result in biased demographic inferences and inaccurate population models.
• Complex models are typically required to account for sampling heterogeneity, increasing analytical challenges.

Purpose of the Study:

• To formally investigate the impact of spatial and temporal sampling heterogeneity on demographic inference.
• To develop a method to circumvent the challenges posed by structured genomic samples in demographic analyses.
• To improve the accuracy of demographic parameter estimation and model selection in population genetics.

Main Methods:

• A new structured approach was integrated into the fastsimcoal2 program.
• The method addresses structured samples without increasing the dimensionality of the site frequency spectrum (SFS).
• The approach was validated using simulated genomic data and modern human genomic datasets.

Main Results:

• The updated fastsimcoal2 program effectively handles spatial and temporal sampling heterogeneity.
• The new SFS-based approach provides more accurate demographic inferences compared to traditional methods.
• The method demonstrates efficiency and relevance in analyzing complex genomic datasets.

Conclusions:

• The developed SFS-based approach offers a robust solution for demographic inference with heterogeneous genomic samples.
• This method is particularly beneficial for analyzing scattered and ancient DNA, crucial for fields like archaeogenetics and conservation genetics.
• The study highlights the importance of accounting for sampling strategies in genomic data analysis to ensure reliable population genetic insights.