Statistical analysis
Two-sided Student’s t-test with unequal variances was done in GraphPad
Prism version 5.01 (GraphPad Software, San Diego, CA, USA). Other
statistical analyses were done in R vers. 4.2.1, using the Brms package
v. 2-18-0 for the Bayesian modelling (Bürkner, 2018). For joint
multilevel modelling of mean colony radiuses and their standard
deviation, the model employing Student’s t likelihood was, in brms model
language, brm(bf(Radius~Strain*temp + (1|day) +
(1|plate), sigma~Strain*temp +
(1|day)+ (1|plate)), data=full, family = student(),
prior= c(prior(normal(0, 5), class=b), prior(normal(0,2), class=”sd”),
prior(normal(0,2), class=”b”, dpar=”sigma”))). For mixture modelling of
mean colony radiuses, the model description is
brm(Radius~Strain*temp + (1|day) +
(1|plate), data=full, family = mixture(gaussian, gaussian)).
The alamarBlue and corresponding OD600 measurements
shown in Fig. 1 were modelled with splines using the blmss package
version 1.1-8 (Umlauf et al. , 2021). The modelling was done
separately for each strain and condition (Alamar signal and
OD600 signal). The model description is bamlss(value
~ s(Time_min), family=”gaussian”). The growth curves in
Fig. 8C were done using the LOESS smoother in the ggplot2 package v.
3.4.0 geom_smooth function (Wickham, 2016; Wickham et al. ,
2019).