I have two datasets, a control and experiment.
Composed of "activation measurements" of cells(y) on fixed time points with uneven intervals (x). At each time point cells from one batch are sampled and analyzed. (control = untreated, experiment = treated)
x1 <- c(0,1,2,5,10,25)
y1 <- c(0, 2, 5,6, 8, 10)
x2 <- c(0,1,2,5,10,25)
y2 <- c(0, 8, 18, 30, 41, 43)
dat1 <- data.frame("x" = x1, "y" = y1, f = "Control")
dat2 <- data.frame("x" = x2, "y" = y2, f = "Experiment")
The activation measurement has a range of (0,100), and the data does show sigmoid curve; which supports the studied measurement (biologically speaking).
I want to show that the trend is different/similar between the control and experiment groups (null = trends are the same)
My approachs (and lost as to what is best):
- fit two trends to a linear model
lm()and compare slopes or coefficients. (link) - fit two trends to nonlinear
nls()andSSlogis()(link) - simple repeated measure anova but assumption of sphericity is confusing for my case(?). (link)
Thoughts/suggestions? Any assistance is appreciated.
