In this publication, we present a statistical approach for estimating the fitness of engineered mutations in high-throughput bulk competitions, which we apply to a data set of the same 560 mutations in yeast Hsp90 under six environmental conditions.

Analyzing the distribution of fitness effects across environments and between one-, two- and three-nucleotide step mutations, we find that larger step sizes harbor more potential for adaptation in challenging environments, but also tend to be more deleterious in the standard environment; an observation consistent with predictions from Fisher’s geometric model (which was proposed by Fisher in 1930).

Furthermore, the shape of the beneficial tail of the fitness distribution becomes heavier under the most severe environmental challenge (here represented by a cold environment with high salinity), indicating a higher potential for rapid adaptation through mutations of unpredictably large size

Link to our paper in Genetics: A Bayesian MCMC approach to assess the complete distribution of fitness effects of new mutations: uncovering the potential for adaptive walks in challenging environments