Hello! I am relatively new to using EdgeR. I am wondering what cpmByGroup() is used for and in what sorts of situations it would be used. I have googled and I don't get many results. Any links to papers or other resources would be helpful.
Thank you so much!
It does what the name says, cpms per group of samples user-defined by group parameter. It's for whatever you need it, heatmaps, boxplots, any downstream analysis that benefits from group-level counts. It's mainly a convenience function, not used in DE testing, see the manual which does not mention it for DE.
Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
version 2.3.6
Thank you so much for the explanation and the quick response! I also noticed that cpmByGroup() isn't mentioned in the EdgeR manual! That is helpful to know that it isn't really used for DE but rather for visualizing. Do you know where I could figure out how cpmByGroup() calculates the average values for groups, and if the function takes into account things like if we ran estimateCommonDisp()?
If there are dispersion estimates in your
DGEList, then they will be used. If not, a dispersion of 0.5 will be used instead. The function usesmglmOneWayto fit a model and extracts the group estimates from that model, which are returned as either CPM or logCPM values, depending on the 'log' argument.