Hey,

I am interested in a set of GO-terms found on Quick-GO. While I found many others, GO:1904936 interneuron migration returns 0 genes. It has many genes in human.

library(biomaRt)
ensembl = useMart("ensembl", dataset="hsapiens_gene_ensembl") #uses human ensembl annotations
 GO="GO:1904936"
(genes <- getBM(attributes=c('hgnc_symbol'), #  'ensembl_transcript_id', 'go_id'
                 filters = "go",  uniqueRows = TRUE,
                 values = GO, mart = ensembl)[,1])

I am a bit clueless here – I guess ensembl does not have all GO-terms that exist on Quick-GO?

Is there any other mart, or mart setting?

I appreciate your input, Thanks, Abel

Try the filter go_parent_term e.g.

library(biomaRt)
ensembl = useEnsembl("ensembl", dataset="hsapiens_gene_ensembl") #uses human ensembl annotations
GO="GO:1904936"
genes <- getBM(attributes=c('hgnc_symbol'), #  'ensembl_transcript_id', 'go_id'
               filters = "go_parent_term",  uniqueRows = TRUE,
               values = GO, 
               mart = ensembl)

> genes
  hgnc_symbol
1         ARX
2       CNTN2
3       FEZF2
4        DRD1
5        LHX6
6        DRD2

That seems to try in pretty well with the QuickGO results I get.

The reason you have to use that filter is because the human genes aren't directly annotated with GO:1904936. Rather they're annotated with GO:0021853 which is a child of your term. Using the go filter only finds things that are explicitly annotated with that term, go_parent_term will get also find those annotated with child terms.