What are the actual differences between different annotation databases?

My lab, for reasons still unknown to me, prefers Ensembl annotations (we're working with transcript/exon expression estimation), while some software ship with RefSeq annotations. Are there significant differences between them today, or are they, for all intents and purposes, interchangeable (e.g., are exon coordinates between RefSeq and Ensembl annotations interchangeable)?


6 Answers 6


To add to rightskewed answer: While it is true that:

Gencode is an additive set of annotation (the manual one done by Havana and an automated one done by Ensembl),

the annotation (GTF) files are quite similar for a few exceptions involving the X chromosome and Y par and additional remarks in the Gencode file (see more at FAQ - Gencode).

What are the actual differences between different annotation databases?

They are a few differences, but the main one for me (and it could be stupid) is

that Refseq is developed by the American NCBI and

the ENSEMBL is mainly developed by the European EMBL-EBI.

Often, labs or people will just start using what is the best known to them (because of a course or workshop) or because they start working with one of the databases with one specific tool and keep with it later.

My lab, for reasons still unknown to me, prefers Ensembl annotations (we're working with transcript/exon expression estimation), while some software ship with RefSeq annotations.

Your lab might be mostly European based people or they might also have read papers like the one from Frankish et al. Comparison of GENCODE and RefSeq gene annotation and the impact of reference geneset on variant effect prediction. BMC Genomics 2015; 16(Suppl 8):S2 - DOI: 10.1186/1471-2164-16-S8-S2

From the Frankish et al. paper paper:

The GENCODE Comprehensive transcripts contain more exons, have greater genomic coverage and capture many more variants than RefSeq in both genome and exome datasets, while the GENCODE Basic set shows a higher degree of concordance with RefSeq and has fewer unique features.

As for:

Are there significant differences between them today, or are they, for all intents and purposes, interchangeable (e.g., are exon coordinates between RefSeq and Ensembl annotations interchangeable)?

No. I don't think they are great differences between them as that the global picture should stay the same (although you will see different results if you are interested in a small set of genes). However, they are not directly interchangeable. Particularly as there are many versions of Ensembl and Refseq based on different genome annotations (and those won't be interchangeable between themselves either in most cases).

However, you can easily translate most[1] of your Refseq IDs to ENSEMBL IDs and vice-versa with tools as http://www.ensembl.org/biomart/martview for example (there are devoted libraries/API as well like Biocondutor: biomaRt

[1] Most as sometimes, they might be annotated in one of the database but haven't (yet) an equivalent in the other.


In fine, even if people tends to keep to what they are used to (and that the annotations are constantly expanded and corrected) depending on the research subject one might be interested in using one database over another:

From Zhao S, Zhang B. A comprehensive evaluation of ensembl, RefSeq, and UCSC annotations in the context of RNA-seq read mapping and gene quantification. BMC Genomics. 2015;16: 97. paper:

When choosing an annotation database, researchers should keep in mind that no database is perfect and some gene annotations might be inaccurate or entirely wrong. [..] Wu et al. [27] suggested that when conducting research that emphasizes reproducible and robust gene expression estimates, a less complex genome annotation, such as RefGene, might be preferred. When conducting more exploratory research, a more complex genome annotation, such as Ensembl, should be chosen.


[27] Wu P-Y, Phan JH, Wang MD. Assessing the impact of human genome annotation choice on RNA-seq expression estimates. BMC Bioinformatics. 2013;14(Suppl 11):S8. doi: 10.1186/1471-2105-14-S11-S8.


Ensembl vs Gencode


The GENCODE annotation is made by merging the Havana manual gene annotation and the Ensembl automated gene annotation. [...] In practical terms, the GENCODE annotation is identical to the Ensembl annotation.

Further, for the GTF file differences:

The only exception is that the genes which are common to the human chromosome X and Y PAR regions can be found twice in the GENCODE GTF, while they are shown only for chromosome X in the Ensembl file.

Gencode(Ensembl) vs RefSeq

Gencode is in almost all cases more comprehensive. For example, this is NCBI RefSeq vs Ensembl (v24, release 83) for BRCA gene: enter image description here

RefSeq and Gencode are not interchangeable in most cases, though RefSeq annotations will often be a subset of the Gencode ones.

  • 10
    $\begingroup$ How is the BRCA screenshot making your point? It's not very obvious at first sight (at least for me) $\endgroup$
    – mgalardini
    May 17, 2017 at 8:45

To add practical advice to what others have said:

In a practical sense, I think the biggest difference between RefSeq and Ensembl/GENCODE is in the sensitivity/specificity trade off.

Ensembl aims more towards the inclusive end, including a far larger number of transcript variants, many of which are only weakly supported.

RefSeq trades some of this sensitivity for specificity - you can be more confident that a RefSeq transcript exists, but less confident that the ReqSeq annotation includes all of the real transcripts for a gene.

  • $\begingroup$ That's why I prefer the Ensembl annotation as you can query for a most confident set by selecting only the Havana (Havana or Ensembl/Havana) transcripts. See: ensembl.org/Help/Faq?id=152 $\endgroup$
    – Mitra
    Oct 16, 2017 at 10:57

While annotations between RefSeq and Gencode are not so different is the coding regions (genes), Gencode is much more rich in the intergenic regions. This could be very advantagous for epigenetic studies, where regulation is of interest.

  • 1
    $\begingroup$ Hi! Is there any publication or other material you could link regarding your claim? That would be very interesting... $\endgroup$
    – mgalardini
    Oct 5, 2017 at 13:54

The UCSC Genome Browser Genes FAQ discusses this question in detail: https://genome.ucsc.edu/FAQ/FAQgenes.html#ens

Officially, the Ensembl and GENCODE gene models are the same. On the latest human and mouse genome assemblies (hg38 and mm10), the identifiers, transcript sequences, and exon coordinates are almost identical between equivalent Ensembl and GENCODE versions (excluding alternative sequences or fix sequences).

GENCODE uses the UCSC convention of prefixing chromosome names with "chr", e.g. "chr1" and "chrM", but Ensembl calls these "1" or "MT". At the time of writing (Ensembl 89), a few transcripts differ due to conversion issues. In addition, around 160 PAR genes are duplicated in GENCODE but only once in Ensembl. The differences affect fewer than 1% of the transcripts. Apart from gene annotation itself, the links to external databases differ.

The GENCODE Release History shows the release dates and can be linked to corresponding Ensembl releases. You can download the gene transcript models from the website https://gencodegenes.org or from http://ensembl.org. For most applications, the files distributed on the GENCODE website should be easier to use, as the third party database links are easier to parse and the sequence identifiers match the UCSC genome files, at least for the primary chromosomes.

Additional information on this question can be found on the GENCODE FAQ page.


Note that the pDot of a certain mutation might differ based on which transcript you are using, and so Ensembl and RefSeq cannot be used interchangeably.

  • $\begingroup$ This is a result of transcript coordinates not aligning 100% (already addressed elsewhere) and does not merit being its own answer. $\endgroup$
    – Ram RS
    Mar 29 at 17:44
  • $\begingroup$ It is a real concern where I work so I thought it would helpful to make it clear to users, but I suppose it is a very niche concern (for example, with the MET gene) $\endgroup$
    – Sarah
    Mar 29 at 18:28
  • $\begingroup$ Yes, the protein coordinates changing can be a problem but the solution is to use MANE transcripts so the annotation is database agnostic. $\endgroup$
    – Ram RS
    Mar 29 at 18:35

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.