Ensembl ID to ENTREZ best converter

Question

[Morris_Chair]

Hello everyone,

I used different tools to convert human gene ensemble ID to ENTREZ but each of them seems to have problems like: -BioMart does not recognizes all the ensemble ID but barely 30% , -BioDBnet ...I couldn't make it working cause I could only select the type of output , -DAVID gives me lots of false duplicates.

What do you usually use for this work ?

thank you

[manaswwm]

I use biomaRt (for ensembl to uniprot id conversion) and this works fine for me (not 100%, I still loose some ids, but this number is small). Maybe you could try to play around a bit with the "attributes" and "filters"?

[ATpoint]

BioMart does not recognizes all the ensemble ID but barely 30%

Please show code and data examples. This is highly unlikely if used properly since biomart directly connects to Ensembl.

[Morris_Chair]

Hi ATpoint , here is the code

mart <- biomaRt::useMart(biomart = "ENSEMBL_MART_ENSEMBL",
                         dataset = "hsapiens_gene_ensembl",
                         host = "http://www.ensembl.org")

genes <- getBM(filters = "ensembl_gene_id",
               attributes = c("ensembl_gene_id","entrezgene_id"),
               values = new_rowname, 
               mart = mart)

Probably there is something wrong with my hosting service?

Thank you

[GenoMax]

Provide examples of identifiers that don't seem to map. It would also help to know where these ID's came from.

[Morris_Chair]

Thank you genomax, indeed.. most of them are non-coding RNA :)

[aaragak1]

BioMart, as others have mentioned, tends to work pretty well. Have you removed the version numbers off the end of the ensembl ID? (ie 1234567.89 --> 1234567)

[Morris_Chair]

Hi aaragak1, yes I did,

thanks

[Morris_Chair]

yes I did that, thanks :)

[russhh]

Are the ensembl IDs that you are working with derived from the current ensembl release?

[Morris_Chair]

Hi russh, yes they are thx, problem solved anyways. Most of those ensemble where from non coding RNA

[Robert.staruch]

Hi Arup Ghosh

Thanks for this code? Is the line Data$entrez calling the in which you want to convert the ENSEMBL Id's? and if so what is row.names(data) referring to - is it saying that the values are the rows in that file data. Should would specify the column in that brackets - such as (data$column).

I am trying to refer to a csv file that I have some ENSEMBL Id's in and I wish to convert these to ENTREZ id's as a new column. I was trying to use the packages above but I get

Error in ENSEMBL(symb_tt3) : could not find function "ENSEMBL"

or

Unknown or uninitialised column: `gene_id`.Error in mapIds_base(x, keys, column, keytype, ..., multiVals = multiVals) : 
  mapIds must have at least one key to match against.

When I use this code:

library("AnnotationDbi") library("org.Hs.eg.db") symb_tt3$gene_id = mapIds(org.Hs.eg.db,
                   keys=ENSEMBL(symb_tt3), 
                    column="ENTREZID",
                    keytype="ENSEMBL",
                    multiVals="first")

[ATpoint]

Please use ADD COMMENT to keep things logically organized.

[Arup Ghosh]

The keys argument is there to specify the column containing the Ensembl ids.

library("AnnotationDbi")
library("org.Hs.eg.db")
symb_tt3$ENTREZID_id = mapIds(org.Hs.eg.db,
                       keys=symb_tt3$ENSEMBL_id_column, 
                        column="ENTREZID",
                        keytype="ENSEMBL",
                        multiVals="first")

This will add a new(ENTREZID_id) column in your dataframe containing the Entrez ids.

Answer 1

[Arup Ghosh]

You can use AnnotationDbi to convert Ensembl ids. The code snippet converts Ensembl id to Entrez ids.

library("AnnotationDbi")
library("org.Hs.eg.db")
#columns(org.Hs.eg.db) # returns list of available keytypes
data$entrez = mapIds(org.Hs.eg.db,
                    keys=row.names(data), #Column containing Ensembl gene ids
                    column="ENTREZID",
                    keytype="ENSEMBL",
                    multiVals="first")

[JJ]

Hi, is this generally acceptable to simply use multiVals="first", I mean how do I know if it is matching to the first entrez ID?

[Arup Ghosh]

The multiVals options default to returning the first occurrence. You can change the option value to list to check for the order of mapped Entrez IDs.

first: This value means that when there are multiple matches only the 1st thing that comes back will be returned. This is the default behavior

Ref: https://bioconductor.org/packages/release/bioc/manuals/AnnotationDbi/man/AnnotationDbi.pdf