Long-term genetic sampling of a plague reservoir reveals a potential selection pressure on Y. pestis to adapt its level of biofilm formation in response to cold and dry climatic conditions. This repository contains the code used to detect regions of the Y. pestis genome likely under selection pressure, the permutation testing code to determine the climatic conditions that coincided or preceded the observed rpoZ variants, and the BEAST XML file by which the Guertu phylogeny was created. * The climate - rpoZ variant analysis has changed substantially compared to version 1. * The description and the workflow of the mutational-hotspots analysis has been made easier to follow compared to version 1 and 2. * In the revised ms we have replaced “mutant” with “variant throughout the text. Mutant was originally referring to sequence variants relative to the reference genome. Although (some of) these variants may be de novo mutations, they could also represent standing variation (minor variants) in the population that are strongly selected for under particular conditions. Thus, we believe that “variants” is the most appropriate term. The source code presented here still uses the old terminology of rpoZ mutant vs rpoZ wildtype. * The output of the file climate.clj is provided as comments within climate.clj (you can also recreate the output by running the file). The comments are missing two lines, which weren't copy-pasted back into the file. Those lines were: ;Frequency with which a random set of 8 non-rpoZ samples has an equal or lower average monthly precipitation than 8 rpoZ samples: 0.03962541656 ;Frequency with which a random set of 8 non-rpoZ samples has an equal or higher average monthly precipitation than 8 rpoZ samples: 0.9603745797