CellChat
CellChat copied to clipboard
jinworks
typo error in `CellChatDB.human` object
I'm using CellChat 2.1.1 and try to investigate the detailed ligand-receptor information. I have noticed that there are several new columns in CellChatDB.human, such as ligand.symbol, and receptor.symbol. However, these newly added columns are not consistent with previous ligand and receptor information.
-
blanks in
ligand.symbolandreceptor.symbol -
typo error
@songmingl It is possible that you see blanks in ligand.symbol or receptor.symbol because we did not carefully check this information. These are only additional information that can help you to inteprete the L-R.
Thank you pointing out these typos. We will look into it later.
@sqjin Thanks!
I have another question, will ligand.symbol or receptor.symbol information be used while calculating probabilities? When plotting gene expression of IL2 signaling, it returns the expression of IL20RA, does this mean that CellChat will extract gene expression through this information?
@songmingl No. The ligand.symbol or receptor.symbol information will be not used in any CellChat's analysis.
You can check the l-r pairs contained in IL2 signaling.
I checked all the interactions with pathway name IL2 in CellChatDB.human$interaction, no IL20RA or IL20RB was found in the records, except for recoptor.symbol. I also checked KEGG: hsa04060, no interaction between ligands shown in following figures and IL20RA/B. I'm confused why IL20RA/B were present in the figure.
By the way, is there a function to extract the gene symbol of each interaction? I want to extract the ligand/receptor genes of some selected pathways and calculate an AUC score for each pathway.
I just checked the source code of plotGeneExpression, the error may be caused by the typo in CellChatDB.human$complex:
@songmingl Thanks. This is indeed a typo in the human database. Will update it later today.
Yes, you can use gene.up <- extractGeneSubsetFromPair(net.up, cellchat).