Fig. 2

Growth temperatures correlate strongly with mean enzyme optima. a The distribution of growth temperatures for all organisms in the growth temperature dataset. b The distribution of experimentally determined optima for the 31,826 enzymes obtained from the BRENDA database. c A Venn diagram showing the number of organisms present in each of the growth temperature dataset, the enzyme temperature dataset, as well as the number of organisms present in both. d Distance score of individual enzyme temperature optima with growth temperature. The horizontal gray line indicates perfect agreement between the two values. The two diagonal dashed lines indicate absolute enzyme temperature optima of 0 °C and 100 °C. The colors indicate the spot density with brighter and more green colors indicating higher density. e A sensitivity plot showing the Pearson correlation coefficient between enzyme temperature optima and growth temperatures. Each data point represents the correlation coefficient obtained between the mean temperature optimum of n sampled enzymes in the range 1 < = n < = 100 and growth temperature. The average of five enzymes or more achieves a correlation coefficient above 0.75. f A scatterplot comparing the growth temperature of organisms with more than five reported enzyme temperature optima with the mean temperature optimum of all enzymes reported for that organism. The diagonal gray line indicates perfect positive correlation. The thick black line represents a locally weighted polynomial regression. The Pearson correlation coefficient is given