A likelihood-based framework for demographic inference from genealogical trees

Area of Science:

• Population genetics
• Computational biology
• Evolutionary genetics

Background:

• Demographic history shapes genetic variation within populations.
• Gene-genealogical trees encode crucial information about population relationships and past events.
• Accurate inference of complex demographic histories is essential for understanding evolutionary processes.

Purpose of the Study:

• To develop a novel framework, genealogical likelihood (gLike), for inferring population demographic history.
• To leverage genealogical information for more accurate demographic inference, particularly in admixed populations.
• To provide a sensitive and accurate method for estimating demographic parameters.

Main Methods:

• Developed a graph-based structure to represent relationships among lineages in gene-genealogical trees.
• Derived the full likelihood across trees under a parameterized demographic model.
• Utilized simulations and empirical data to validate the gLike framework.

Main Results:

• gLike accurately estimates numerous demographic parameters, including ancestral population sizes, admixture timing, and proportions.
• The method demonstrates superior performance compared to conventional site frequency spectrum-based methods for admixed populations.
• gLike shows high sensitivity and accuracy in inferring complex demographic histories.

Conclusions:

• The genealogical likelihood (gLike) framework offers a powerful new tool for demographic inference.
• gLike effectively harnesses genealogical information, providing deeper insights into population histories.
• This method has broad applicability for studying the demography of diverse species, including humans.