I know that with GRCh38, it is possible to extract the matching RefSeq transcripts for Ensembl genes/transcripts/proteins using Biomart from Ensembl.

However, for GRCh37, this option does not exists. Using the following MySQL command returns Ensembl transcript IDs for matching RefSeq transcripts:

mysql -u anonymous -h ensembldb.ensembl.org -A -P 3337 -D homo_sapiens_core_99_37 -e "SELECT transcript.stable_id, xref.display_label FROM transcript, object_xref, xref,external_db WHERE transcript.transcript_id = object_xref.ensembl_id AND object_xref.ensembl_object_type = 'Transcript' AND object_xref.xref_id = xref.xref_id AND xref.external_db_id = external_db.external_db_id AND external_db.db_name = 'RefSeq_mRNA';"

Also, I was trying to play around with the databases schema but I am not able to find any solution to problem.

Ultimately, I wish to obtain a table like the following:

Gene stable ID      Transcript stable ID    Protein stable ID   RefSeq match transcript
ENSG00000120329     ENST00000239451         ENSP00000239451     NM_031947.4
ENSG00000164219     ENST00000419445         ENSP00000404676     NM_005023.4
ENSG00000150594     ENST00000280155         ENSP00000280155     NM_000681.4
ENSG00000113578     ENST00000337706         ENSP00000338548     NM_000800.5

Thank you in advance.


2 Answers 2


The RefSeq match option in BioMart is from the Matched Annotation from NCBI and EBI (MANE) collaboration between RefSeq and Ensembl. It has only been calculated for the up-to-date gene annotation on GRCh38 so cannot be obtained on GRCh37. You can get mapping from Ensembl to RefSeq transcripts through BioMart as RefSeq mRNA ID (refseq_mrna in R) but this is not a perfect match like the MANE, it is a mapping based on sequence similarity and similar genomic location, and there can be mismatches between them.

  • $\begingroup$ Thank you @Emily_Ensembl for your answer. I will continue discussing this topic here. Is it possible to obtain also the RefSeq mRNA ID with the versions using biomaRt ? $\endgroup$
    – user324810
    Commented Jan 21, 2020 at 22:31
  • $\begingroup$ No, we don't have the versions on the old verion of the database. $\endgroup$ Commented Jan 22, 2020 at 9:46

Try the GRCh37 version of BioMart, and the External reference "RefSeq mRNA ID" / "RefSeq ncRNA ID":


From the Perl code provided by BioMart, it looks like the field name should be "refseq_mrna" or "refseq_ncrna". I'm not sure how that translates to the SQL code.

I feel compelled to add a caution to this, though. GRCh37 is a very old build, having been replaced by GRCh38 in December 2013. If possible, please find a way to use GRCh38 for your analysis instead, and don't create any new software that depends on GRCh37. For example, gnomAD v3 supports GRCh38 (see bioconductor, or here).

  • $\begingroup$ Thank you for your comment. Unfortunately, I have to use GRCh37 since gnomAD were annotated with this old version using VEP from Ensembl. So, doing a lift-over or annotating again would lose any supports that the paper had at that time. $\endgroup$
    – user324810
    Commented Jan 22, 2020 at 10:23

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.