Thanks for all your work. I'm currently following the CATALYST workflow and ran into a question when I was trying to get the createContrast() for the diffcyt-DA-GLMM to work. In my 'condition' parameter I have 3 levels: Healthy, Non-Hospitalized and Hospitalized. As far as I understand createContrast(c(0,1,0)) compares Non-Hospitalized to the Healthy group and createContrast(c(0,-1,1)) compares the Non-Hospitalized to the Hospitalized group. But what output do I get when I use createContrast(c(0,1,1))? I'm trying to get p-values for comparisons between each of the three groups (Healthy vs. (Non-)Hospitalized and Non-hospitalized vs. Hospitalized). The only way I can currently see to make that happen is to run the analysis multiple times, each time with a different contrast matrix. I'm wondering if that is statistically allowed though (I'm not a statistician) and if not, if there is a way to do this properly.
Hi @lmweber,
Thanks for all your work. I'm currently following the CATALYST workflow and ran into a question when I was trying to get the
createContrast()
for the diffcyt-DA-GLMM to work. In my 'condition' parameter I have 3 levels: Healthy, Non-Hospitalized and Hospitalized. As far as I understandcreateContrast(c(0,1,0))
compares Non-Hospitalized to the Healthy group andcreateContrast(c(0,-1,1))
compares the Non-Hospitalized to the Hospitalized group. But what output do I get when I usecreateContrast(c(0,1,1))
? I'm trying to get p-values for comparisons between each of the three groups (Healthy vs. (Non-)Hospitalized and Non-hospitalized vs. Hospitalized). The only way I can currently see to make that happen is to run the analysis multiple times, each time with a different contrast matrix. I'm wondering if that is statistically allowed though (I'm not a statistician) and if not, if there is a way to do this properly.Thanks!
Martijn