tumr is a collection of tools for analyzing tumor growth data.
An example of how to use tumr can be seen in the Get Started page.
tumr includes:
tumr() - creates a tumr object that can be passed into plot_median(), rfeat(), and lmm().
plot_median() - creates a spaghetti plot and plots the median tumor growth measurement for each treatment group.
-
rfeat() - tests whether the beta coefficients (slope of tumor measure over time) of individuals in different treatment groups are different from each other using either t-test, ANOVA, Tukey, or both ANOVA and Tukey.
- rfeat() has a plot() function that shows the individual beta coefficients and mean slope by treatment group.
-
lmm() - fits a linear mixed model with the default formula of log1p(Volume) ~ Treatment * scale(Day) + (scale(Day) | ID) where the time variable is centered.
- lmm() includes a summary() function that provides the overall effect of time averaged over the levels of the treatment group, the slope of treatment over time for each treatment group, and the results of a contrast to test the slope differences between the treatment groups.
- lmm() includes a plot() function that shows the predicted values of the tumor growth measurements over time and the mean slope for each treatment with their 95% confidence intervals.
-
bhm() - fits a Bayesian Hierarchical Linear Model.
- bhm() includes a summary() function that provides the treatment-specific intercepts with exponential transformation , treatment-specific slopes with exponential transformation, and pairwise treatment contrasts in slopes with exponential transformation.
- bhm() includes a plot() function that shows the the mean slope for each treatment with their 90% credible intervals, Pairwise slope contrasts with their 90% credible intervals, and MCMC trace plots for model diagnostics.
How to install tumr
To install tumr, copy and paste the following code into the console
if (!requireNamespace("remotes")) install.packages("remotes")
remotes::install_github("pbreheny/tumr")