The role of weak selection and high mutation rates in nearly neutral evolution

Daniel J. Lawson, Henrik Jeldtoft JensenJournal of Theoretical Biology 2009 · DOI

TL;DR

This paper investigates how weak selection and high mutation rates can lead to "effectively neutral" evolution, even when fitness differences exist between types. The authors introduce a model that explains the prevalence of neutral evolution in large populations, particularly in asexual organisms, by demonstrating the formation of "neutral networks" where slightly less fit types can still persist.

What Problem Does This Paper Address?

Evolutionary theory traditionally grapples with the balance between selection (driven by fitness differences) and random drift (chance events). While the "neutral theory of molecular evolution" posits that many evolutionary changes are due to drift, empirical observations in large populations, especially asexual ones like bacteria and viruses, often show patterns that seem neutral despite the potential for selective differences. This paper addresses the question of how neutral dynamics can arise and persist in such scenarios, particularly when mutation rates are high and selection is weak, challenging the assumption that neutral evolution requires near-identical fitness across all types.

Methods

The authors develop a theoretical model based on a Moran birth/death process for a haploid population with a conserved number of individuals. They explore three types of fitness landscapes:

  • Random uncorrelated fitness landscape: A highly rugged landscape where fitness values are assigned randomly and independently.
  • Truncation fitness landscape: A "top hat" function where fitness drops sharply outside a specific range of type space.
  • Linear fitness landscape: Fitness increases linearly with type, representing directional selection.

They use a combination of analytical "mean-field" arguments and numerical simulations to analyze the conditions under which effectively neutral dynamics are observed. A key analytical tool is the "data collapse" method, which involves transforming data from different population sizes to a single functional form to identify scaling relationships.

Datasets

Not discussed in the paper. The study relies on simulated data generated by the authors' model.

Key Insights and Ideas

  • Neutral Networks: The paper demonstrates that even with fitness differences, "effectively neutral" networks of types can form. These networks connect types that are nearly equally fit, allowing for neutral drift to dominate within the network.
  • High Mutation Rates are Key: For large populations, high mutation rates are crucial for maintaining a distribution of types, which in turn facilitates the formation of these neutral networks. This contrasts with low mutation rate models where only a few types typically compete.
  • "Effectively Neutral" is Context-Dependent: The definition of "effectively neutral" depends on population size and mutation rate. Weak selection can be rendered irrelevant by high mutation rates and large population sizes, leading to neutral dynamics.
  • Asexual Organisms are Prime Candidates: The model's findings are particularly relevant to asexual populations (like bacteria) where recombination is limited, making them more susceptible to the dynamics described.
  • Justification for Neutral Null Models: The work provides a theoretical basis for using neutral models as a null hypothesis in evolutionary studies, even when some level of selection is present.

Key Results

  • The critical selection strength ($s^*$) for observing neutral dynamics scales with population size ($N$) and dimension ($D$) in specific ways, e.g., $s^* \propto N^{-\alpha(D)}$. The exponent $\alpha(D)$ depends on the fitness landscape.
  • For the random fitness landscape, $\alpha(D)$ decreases with dimension, approaching 0 for large $D$.
  • For the truncation and linear landscapes, $\alpha(D)$ is constant (1 and 3/2, respectively).
  • The time-averaged variation ($\langle w \rangle$) and genetic distance measures are robust indicators of neutral dynamics and show a consistent critical selection strength ($s^*$) across different indicators.
  • The model predicts that neutral dynamics can persist for larger selection strengths and population sizes than previously thought, especially in asexual populations.

Limitations and Future Work

  • Static Fitness Landscapes: The models use static fitness landscapes, which may not fully capture the dynamic and environmentally dependent nature of real-world fitness.
  • Computational Demands: Calculating genetic distance measures for large populations and dimensions is computationally intensive, limiting the tested parameter space.
  • Applicability to Genetics Data: While conceptually relevant, the model is not directly applicable to inferring population size from genetic data alone.
  • Future Work: The authors suggest exploring more complex fitness landscapes, investigating the role of recombination, and further integrating these findings into existing theoretical frameworks for evolutionary biology. They also highlight the need for deeper understanding of the relationship between their model and other approaches, such as those focusing on selection for mutational robustness.

Summary updated

BibTeX
@article{lawson2009role,
  author = {Daniel J. Lawson and Henrik Jeldtoft Jensen},
  title = {{The role of weak selection and high mutation rates in nearly neutral evolution}},
  journal = {Journal of Theoretical Biology},
  year = {2009},
  doi = {10.1016/j.jtbi.2008.12.029},
}
Figure