Dear all, I have faced with a problem in using clusterProfiler.
I have 152 differential expressed genes between cancer and normal tissues and I want to enrich their biological processes using clusterProfiler.
I was recommended to use enrichGO function and limit the background to the list of genes that have been expressed in both cancer and normal tissues. So I filtered genes with less than 10 counts and convert the ensembl IDs to gene symbol using biomaRt package. The same conversion was performed for 152 DEGs.
However when I run clusterProfiler I got the message than "no gene can be mapped" and that "it expected IDs like: TEKT3,MARF1,...".
I suspect to my inputs, so I compare the list of 152 DEGs with the list of backgrounds which consist of 12604 gene symbol. Strangely small number of DEGs were present in the background.
Is it possible that biomaRt had made mistake in ID conversion?
I will appreciate any recommendation
Nazanin
Try converting your ids to entrez ids as in the clusterProfiler manual here using
bitr
function.I think this tools doesn't work with gene symbols.
no it does work with gene symbol if you specify the
keyType = "SYMBOL"
I used SYMBOL for both my list and background at the first step ,however I got the message that no gene can be mapped. My list include 152 gene and the background is around 12000 genes. 30 out of 152 genes were present in the background.
see https://guangchuangyu.github.io/2016/01/go-analysis-using-clusterprofiler/ and the package vignette.
You need to specify the ID type if it is supported.