Coalescence and translation: A language model for population genetics.

Clinical Snapshot

Key Findings

• P Value: Not applicable for methodological study

• Effect Size: Matching accuracy in-distribution, approaching accuracy out-of-distribution

• Primary Outcome: Competitive accuracy with MCMC-based methods across demographic scenarios

• Nnt Or Sensitivity: Over one million coalescence predictions achievable in minutes vs hours-days for traditional methods

• Confidence Interval: Well-calibrated approximate posteriors provided

Clinical Application

High feasibility for research applications with appropriate computational infrastructure Highly relevant for Australian population genetic research including Indigenous genomics, conservation genetics, and agricultural genomics research supported by Australian research institutions Researchers conducting population genetic analyses, particularly large-scale genomic studies

Abstract

Probabilistic models such as the sequentially Markovian coalescent have long provided a powerful framework for population genetic inference, enabling reconstruction of demographic history and ancestral relationships from genomic data. However, these methods are inherently specialized, relying on predefined assumptions and/or limited scalability. Recent advances in simulation and deep learning provide an alternative approach: learning directly to generalize from synthetic genetic data to infer hidden evolutionary processes. Here we reframe the inference of coalescence times as a problem of translation between two biological languages: the sparse, observable patterns of mutation along the genome and the unobservable ancestral recombination graph that gave rise to them. Inspired by large language models, we develop cxt, a decoder-only transformer that autoregressively predicts coalescent events conditioned on local mutational context. We show that cxt performs competitively with state-of-the-art Markov Chain Monte Carlo-based likelihood models across a broad range of demographic scenarios, matching their accuracy in-distribution and approaching it in out-of-distribution, with the potential for improvement via fine-tuning. Trained on simulations spanning the stdpopsim catalog [J. R. Adrion et al., eLife 9, e54967 (2020); M. E. Lauterbur et al., eLife 12, e84874 (2023); G. Gower et al., Accessible, realistic genome simulation with selection using stdpopsim. bioRxiv [Preprint] (2025). https://doi.org/10.1101/2025.03.23.644823.], the model generalizes robustly and enables efficient inference at scale, producing over a million coalescence predictions in minutes. cxt also produces well-calibrated approximate posteriors, enabling principled uncertainty quantification. We apply cxt to population genomic data from both humans and mosquitoes, highlighting the model's ability to deal with the complexities of empirical data.

References

1. 1.Adrion, J. R., Cole, C. B., Dukler, N., Galloway, J. G., Gladstein, A. L., Gower, G., Kyriazis, C. C., Ragsdale, A. P., Tsambos, G., Baumdicker, F., Carlson, J., Cartwright, R. A., Durvasula, A., Gronau, I., Kim, B. Y., McKenzie, P., Messer, P. W., Noskova, E., Ortega-Del Vecchyo, D., ... Kern, A. D. (2020). A community-maintained standard library of population genetic models. eLife, 9, e54967. https://doi.org/10.7554/eLife.54967
2. 2.Lauterbur, M. E., Cavassim, M. I. A., Gladstein, A. L., Gower, G., Pope, N. S., Tsambos, G., Adrion, J., Belsare, S., Biddanda, A., Caudill, V., Cury, J., Echevarria, I., Haller, B. C., Hasan, A. R., Huang, X., Iasi, L. N. M., Noskova, E., Obsteter, J., Pavinato, V. A. C., ... Kern, A. D. (2023). Expanding the stdpopsim species catalog, and lessons learned for realistic genome simulations. eLife, 12, e84874. https://doi.org/10.7554/eLife.84874