Accelerated simulation of evolutionary trajectories in origin-fixation models.

Area of Science:

• Evolutionary Biology
• Computational Biology
• Population Genetics

Background:

• Origin-fixation models are crucial for understanding evolutionary dynamics.
• Traditional simulation methods can be computationally intensive, limiting their application with complex fitness functions.
• The effective population size (Ne) heavily influences the computational cost of existing algorithms.

Purpose of the Study:

• To introduce an accelerated algorithm for forward-simulation of origin-fixation models.
• To reduce the number of fitness evaluations required per fixed mutation.
• To provide a computationally efficient yet accurate simulation tool for evolutionary biology.

Main Methods:

• Developed an accelerated algorithm for origin-fixation simulations.
• Compared the accelerated algorithm against traditional methods using evolutionary metrics.
• Analyzed the distribution of fixed selection coefficients and probability of reversion.
• Investigated variance in selection coefficients and population size rescaling.

Main Results:

• The accelerated algorithm requires approximately two fitness evaluations per fixed mutation, a significant reduction from traditional methods (order of Ne).
• The accelerated algorithm produces the same steady-state results as the original algorithm but alters the order of fixed mutations.
• While generally equivalent, the accelerated algorithm shows less variance in fixed selection coefficients, which can be recovered by rescaling population size.
• A linear relationship was found between rescaled and original population sizes.

Conclusions:

• The accelerated algorithm offers a computationally efficient alternative for origin-fixation simulations.
• This method enhances the feasibility of simulating complex fitness functions in evolutionary biology.
• The algorithm provides a valuable tool for increasing computational complexity without substantial loss of simulation accuracy.