We introduce BESTish, an exact Bayesian method to infer mutation rates and selection coefficients from variant allele frequencies.

We introduce BESTish, a novel parameter estimation method utilizing explicit mean-field dynamics and Gaussian-Markov approximations to infer mutation rates and selection coefficients for drivers of clonal hematopoiesis. BESTish is available on Github, and can be applied to data from either population-level or longitudinal studies, with potential applications in other biological contexts.