Inferring population history from genealogical trees

Area of Science:

• Population genetics
• Molecular evolution
• Bioinformatics

Background:

• Inferring population history from DNA sequence data is crucial for understanding human and viral evolution.
• Key questions involve detecting population expansions and their timing, relevant for both human demographics and viral epidemiology (e.g., HIV-1, Hepatitis C).

Purpose of the Study:

• To investigate the accuracy of inferring population history using single-locus DNA data.
• To determine if population history parameters can be reliably estimated from DNA sequence samples.

Main Methods:

• Theoretical analysis based on Kingman's coalescent process.
• Consideration of an idealized scenario observing the gene tree for 'n' non-recombining, selectively neutral DNA sequences.
• Examination of the theoretical upper limit of information extractable from a sample.

Main Results:

• Consistent estimation of population history parameters (e.g., growth rate) is not achievable as sample size ('n') increases.
• This limitation is more severe than typically observed for genetic parameter estimators, such as mutation rate convergence.
• The study presents findings on the distribution of maximum likelihood estimators for population history.

Conclusions:

• Accurate inference of population history from single-locus DNA data is fundamentally limited, particularly with larger sample sizes.
• The coalescent process highlights inherent challenges in reconstructing past population dynamics solely from sequence data.