Inferring Continuous and Discrete Population Genetic Structure Across Space

Area of Science:

• Population Genetics
• Evolutionary Biology
• Bioinformatics

Background:

• Characterizing population structure is challenging due to continuous genetic differentiation and discrete barriers.
• Current methods may misinterpret continuous genetic processes as discrete population structures, especially with discontinuous sampling.
• Existing approaches struggle to visualize population structure in geographically distributed populations.

Purpose of the Study:

• To develop a statistical framework for simultaneously inferring continuous and discrete patterns of population structure.
• To address the "clines versus clusters" problem in population genetics.
• To improve the inference and visualization of population structure in genetic data.

Main Methods:

• A novel statistical framework is presented for inferring population structure.
• The method estimates ancestry proportions from two-dimensional population layers.
• It models the decay of genetic relatedness with distance within each layer.

Main Results:

• The framework explicitly distinguishes between continuous "clines" and discrete "clusters" in genetic variation.
• It offers a remedy for overfitting issues common in nonspatial population structure models.
• The method provides effective descriptions of genetic relatedness in interacting, geographically distributed populations.

Conclusions:

• The developed statistical framework accurately characterizes population structure by integrating continuous and discrete patterns.
• This approach is valuable for studying populations experiencing range expansions or secondary contact.
• The method's utility is validated through simulations and real-world datasets of poplars and black bears.