I am receiving an error from Clusterprofiler when calling wikipathway functions. The error message I receive has been discussed in previous posts (Clusterprofiler error: No gene can be mapped) and on the Clusterprofile vignette FAQs section.
x <- read.table("/Users/matthew/Bioinformatics/cytoNuc/counts_table_res_cyto_Cyto_vs_Nuc.txt",
header=TRUE)
x.bkgd.genes <- x[,1]
x.bkgd.genes.entrez <- bitr(x.bkgd.genes,fromType = "SYMBOL",toType = "ENTREZID",OrgDb =
org.Hs.eg.db)
x.entrez <- merge(x, x.bkgd.genes.entrez, by.x = "row", by.y = "SYMBOL")
geneList <- x.entrez[,3]#LFC values
names(geneList) <- as.character(x.entrez[,8]) #entrez ids of my genes
geneList = sort(geneList, decreasing = TRUE)
head(geneList)
652 100133941 100129792 389114 4685 2625
10.001578 9.789602 9.472782 9.212101 8.918846 8.429205
This gene list works for other Clusterprofiler functions like gseKEGG and enrichGO, yet for the two functions shown below the error is produced. The error messages that have previously been discussed include some output following "Expected input gene ID:" but I do not have anything following. Previous post on this issue have attributed the error to using mouse entrez IDs mistakenly which is not accepted in Clusterprofiler, but I am working with human IDs only.
gene <- names(geneList)[abs(geneList) > 2]
enrichWP(gene, organism = "Homo sapiens")
and
gseWP(geneList, organism = "Homo sapiens")
produce
--> No gene can be mapped....
--> Expected input gene ID:
--> return NULL...
Please let me know if you have any suggestions.
From what I remember, wikipathways is quite a bit smaller than KEGG or GO databases. Your problem might be explained if your gene list is small and none of your genes have any annotated function in wikipathways.