Viral Simulation Reveals Overestimation Bias in Within-Host Phylodynamic Migration Rate Estimates Under Selection

Area of Science:

• Virology
• Evolutionary Biology
• Computational Biology

Background:

• Phylodynamic methods infer viral population dynamics within hosts.
• These methods are crucial for understanding HIV-1 transmission between anatomical compartments.
• Current methods often assume neutral evolution, ignoring genomic selection pressures.

Purpose of the Study:

• To investigate the impact of selection on phylodynamic migration rate estimates.
• To assess biases in migration rate estimations under different evolutionary models.
• To evaluate the accuracy of phylogeographic methods in selective environments.

Main Methods:

• Developed 'virolution,' an agent-based simulation tool for within-host virus evolution.
• Simulated viral sequences and genealogies under neutral and purifying selection.
• Applied stochastic mixture and structured coalescent models in BEAST2.

Main Results:

• Purifying selection significantly overestimates migration rates.
• Bias was observed in both stochastic mixture and structured coalescent models.
• Common phylogeographic methods assuming neutrality are unreliable under selection.

Conclusions:

• Selection significantly biases phylodynamic migration rate estimates.
• Current phylogeographic methods require re-evaluation for use in selective regimes.
• Findings have critical implications for HIV-1 transmission dynamics research.