03 June, 2008

Ensembl 11 - UCSC 8

The past days I was in Barcelona at the European Human Genetics Conference 2008. After giving my presentation on Ensembl in one of the 'Educational sessions' and listening to numerous talks about GWAS (genome-wide association studies), I had a look at the posters. Under the impression that the UCSC Genome Browser is the preferred browser amongst (human) geneticists and with Ewan's experience at the recent 'Biology of Genomes' meeting fresh in my mind, I decided to have a closer look at the posters in the Cytogenetics section. Out of 189 posters, I could positively identify 11 with Ensembl screenshots (mostly CytoView and ContigView, but also two times KaryoView), 8 with UCSC Genome Browser screenshots and none with NCBI Map Viewer screenshots. OK, I admit that I can recognise almost any pixel copied from our site and may have missed one or two UCSC screenshots, but all in all I thought this was a very encouraging result! Of course we should keep in mind that this was a European conference, mainly attended by European scientists .... I guess I have a bit more screenshot counting to do at the International Congress of Genetics 2008 in Berlin. So, let's say Ensembl 11 - UCSC 8 is the score at half time .... next month I'll report back with the final result!


Rileen said...

That's interesting - as a Bioinformatics Ph.D student in an institute of mainly Biologists and Biochemists (http://www.fli-leibniz.de/), I've seen both sides, so to say.

My group (Genome Analysis) is biased towards the UCSC site, partly because you can "see a lot at a glance", and partly because the tables make life easier for those of us who do any coding or scripting.

On the other hand, collaborators from other groups as well as other institutes seem to provide Ensembl IDs for genes of interest more often than not.

Personally I tend to use the UCSC browser more (with the Ensembl genes track turned on!), but if I find anything interesting, I often cross check with Ensembl. That to me is the best thing about having two major browsers rather than just one.

I'll look forward to the "final score" :-)

Bert (Helpdesk team) said...

Hello Rileen,

That is also what we saw in a user survey that we did last year; quite some people seem to use both browsers.

In the workshops I give people sometimes say that it would make their life much easier if there was only one genome browser. Personally I think, however, that having two browsers is a good thing, as some -friendly(!!!)- competition makes, in my humble opinion, both browsers better, which is in the interest of all users ....


Rileen said...

I absolutely agree - having multiple browsers is better for the browsers as well as the users.

An unrelated question - do you have some email address on which the Vega browser team/helpdesk can be reached? I want to ask some questions, but the hyperlink for feedback actually directs one to the help page.

If you happen to be responsible for Vega as well, here's one question :-) :

The stats show 559,017 "Annotated exons" in human, which seems really, really high. Does this come about because exons are actually counted multiple times (i.e once per transcript that they appear in), or are there actually that many unique, annotated exons?

Ensembl Helpdesk (Giulietta) said...

Hi Rileen,

If you click on the 'Contact Helpdesk' button at the top right of the Vega 'help' page, it will direct you to the appropriate form. Or, email: vega@sanger.ac.uk

Ensembl and Vega are different projects, see a comparison here:


Vega relies on manual annotation, rather than an automatic annotation pipeline, as found in Ensembl. The Vega website is powered on Ensembl software, so it looks similar. Also, Ensembl imports genes from Vega/Havana for comparison.

To be sure, do send Vega an email, but the large number of exons will most likely reflect pseudogenes, along with splice isoforms and protein coding transcripts.

Giulietta (Ensembl Helpdesk)

Rileen said...


Thanks for that. I am quite aware that Vega is a different project - in fact, my group was one of the contributors (back when the institute used to be the Institute of Molecular Biotechnology, Jena).

And being that Vega is manually annotated, it was all the more surprising to see such a high number. It is too high to be accounted for by splice variants, unless exons are counted multiple times, but I have no idea how much might be contributed by pseudogenes.

Anyway, I shall mail Vega. I asked via a comment here because I knew that the Ensembl and Vega teams collaborate,not because I thought they are the same project. Thanks once again for the email address.