The Effects of Population Size Histories on Estimates of Selection Coefficients from Time-Series Genetic Data

Abstract

Many approaches have been developed for inferring selection coefficients from time series data while accounting for genetic drift. These approaches have been motivated by the intuition that properly accounting for the population size history can significantly improve estimates of selective strengths. However, the improvement in inference accuracy that can be attained by modeling drift has not been characterized. Here, by comparing maximum likelihood estimates of selection coefficients that account for the true population size history with estimates that ignore drift by assuming allele frequencies evolve deterministically in a population of infinite size, we address the following questions: how much can modeling the population size history improve estimates of selection coefficients? How much can mis-inferred population sizes hurt inferences of selection coefficients? We conduct our analysis under the discrete Wright–Fisher model by deriving the exact probability of an allele frequency trajectory in a population of time-varying size and we replicate our results under the diffusion model. For both models, we find that ignoring drift leads to estimates of selection coefficients that are nearly as accurate as estimates that account for the true population history, even when population sizes are small and drift is high. This result is of interest because inference methods that ignore drift are widely used in evolutionary studies and can be many orders of magnitude faster than methods that account for population sizes.

Detecting and Quantifying Natural Selection at Two Linked Loci from Time Series Data of Allele Frequencies with Forward-in-Time Simulations

“…In practice, it will be necessary to find a good approximation of the Wright-Fisher model for the method to be computationally feasible, which will be the topic of future investigation. An important consideration is to what degree the results of the inference of natural selection are affected by the choice of stochastic or deterministic dynamics for the allele frequency trajectories (Jewett et al, 2016), and whether in many scenarios approximation with a deterministic model is satisfactory.…”

Assessing the Relationship of Ancient and Modern Populations

“…Advances in DNA isolation ( Dabney et al 2013 ), library preparation ( Meyer et al 2012 ), bone sampling ( Pinhasi et al 2015 ), and sequence capture ( Haak et al 2015 ) make it possible to obtain genome-wide data from hundreds of samples ( Allentoft et al 2015 ; Haak et al 2015 ; Mathieson et al 2015 ; Fu et al 2016 ). Analysis of these data can provide new insight into recent evolutionary processes, which leave faint signatures in modern genomes, including natural selection ( Jewett et al 2016 ; Schraiber et al 2016 ) and population replacement ( Lazaridis et al 2014 ; Sjödin et al 2014 ).…”

Inferring Fitness Effects from Time-Resolved Sequence Data with a Delay-Deterministic Model

“…Deterministic models of adaptation have been proposed as a rapid and effective method for inferring selection from time-resolved sequence data ( Jewett et al 2016 ). Here, we have highlighted a limitation of such frameworks whereby a deterministic model may severely underestimate the magnitude of selection in a system.…”

“…Numerical solution of the stochastic dynamics of the population may be computationally intensive, inspiring the development of more rapid propagation methods and the consideration of potential alternative solutions ( Khatri 2016 ; Krukov et al 2017 ; Nené et al 2018 ). In a recent work, considering a range of potential models for the demographic history of a population, it was concluded that deterministic approximations to evolution under drift can produce accurate estimates of the magnitude of selection ( Jewett et al 2016 ). Such models of selection, mutation, and recombination have been used to generate insights into viral adaptation ( Ganusov et al 2011 ; Illingworth 2015 ; Sobel Leonard et al 2017 ).…”