Accelerated Bayesian inference of population size history from recombining sequence data

Area of Science:

• Population genetics
• Computational biology
• Genomic data analysis

Background:

• Inferring population history is crucial for understanding evolutionary processes.
• Existing methods for demographic inference from genomic data have limitations in speed and accuracy.
• Accurate reconstruction of past population events requires robust computational tools.

Purpose of the Study:

• Introduce Population History Learning by Averaging Sampled Histories (PHLASH), a novel method for inferring population history.
• Develop an accurate and adaptive estimator for demographic inference using whole-genome sequence data.
• Provide a computationally efficient tool for population genetics research.

Main Methods:

• PHLASH utilizes random, low-dimensional projections of the coalescent intensity function.
• It averages these projections from the posterior distribution of a coalescent-like model.
• A key advance is a new algorithm for computing the score function of coalescent hidden Markov models.

Main Results:

• PHLASH demonstrates faster performance and lower error rates compared to SMC++, MSMC2, and FITCOAL on simulated data.
• The method offers automatic uncertainty quantification.
• It enables new Bayesian testing procedures for population structure and ancient bottlenecks.

Conclusions:

• PHLASH is an accurate, adaptive, and computationally efficient method for inferring population history from genomic data.
• The software package is user-friendly and leverages GPU acceleration.
• PHLASH advances the field of population genetics by providing improved tools for demographic inference and hypothesis testing.