Population-scale Ancestral Recombination Graphs with tskit 1.0

Citation (APA 7)

Jeffery, B., Wong, Y., Thornton, K., Tsambos, G., Bisschop, G., Deng, Y., Ellerman, E. C., Forest, T. B., Fritze, H., Goldstein, D., Gorjanc, G., Gower, G., Gravel, S., Guez, J., Haller, B. C., Kern, A. D., Kirk, L., Lee, H., Lehmann, B., … Kelleher, J. (2026). Population-scale ancestral recombination graphs with tskit 1.0. Nature Methods. https://doi.org/10.1038/s41592-026-03325-1

Abstract

Ancestral recombination graphs (ARGs) capture the full genetic history of samples from a recombining species. Although ARGs have been a central theoretical object in population genetics for decades, their practical use was constrained by the lack of scalable inference methods, standard interchange formats and software infrastructure. Recent breakthroughs in simulation and inference have substantially changed this landscape, which has led to renewed interest in ARG-based analyses across population and statistical genetics. The tskit library has had a key enabling role in this shift and has become foundational infrastructure for working with ARGs. This Correspondence marks the release of tskit 1.0, which formalizes long-term stability guarantees for its data formats and application programming interfaces (APIs).