[ome-users] number of institutions using OMERO

Sun Jun 26 15:36:46 BST 2011

Hi Brian-

We've had these requests before.  It seems that most people prefer answering these questions directly, and not on list.  We have a number of letters of support on file from people using OMERO in production, so they do really exist!  ;-)!  Maybe they will write to you directly.  

I'll point you to a few public resources that might help and describe our own installation.

First, a few examples of OMERO in production:

-- The DMBI at JIC:  http://dmbi.nbi.bbsrc.ac.uk/index.php/Main_Page

-- There are a number of presentations describing use of OMERO from our most recent OME Users Meeting:

http://www.openmicroscopy.org/site/community/minutes/meetings/june-2011-paris-users-meeting/presentations

-- Slightly derivative-- two on-line image data repositories that use OMERO:  

http://jcb-dataviewer.rupress.org/
http://www.celllibrary.org/

And of course numerous installations of Columbus®, from Perkin-Elmer (http://www.perkinelmer.co.uk/pages/020/cellularimaging/products/columbus.xhtml).

At Dundee our production OMERO server is  a dual quad core server, with 8 GB RAM, fibre-channeled into 10 TB of disk space.  The OMERO instance uses the College LDAP system, so everyone has an account automatically.  Data is imported from a central 220 TB data store, which is shared across all users in the College.  Because of limited disk we haven't advertised our OMERO instance as a central resource, but word has gotten out, especially about a series of Matlab routines we run for very standard image segmentation, so a number of people outside my own lab use it (the Matlab resource runs on a separate Windows server, and connect to the OMERO server).  Because of the limited disk space on our OMERO installation (which we are fixing as part of our new grant), some experiments that generate very large datasets (>50 GB) don't use this facility.    This will change dramatically as we start enabling OMERO.fs.

The OME Team is starting to see many requests for the kind of thing you are asking:  "... manage image data across multiple centers/institutes and research labs".  We actually run accounts for Dundee and a few other sites from our own production server-- as party of collaborations, etc.  This of course depends on good network backbone connectivity (this is good in the UK) and good connectivity from your site to the backbone. Taking the view that keeping things simple pays in the end, it's probably a good idea to use a single installation, properly spec'd for the usage you expect to receive, and put data there as much as possible.  We all know that this can't satisfy all needs, so any detail regarding how you'd like distributed resources to work would be most helpful.

We (my lab) can run our installation because we (our College) has invested heavily into informatics infrastructure.  We have a great network, a reasonable data centre, high performance storage and backup, and good people running these and other services (LDAP, etc.).  This means that our OMERO instance can benefit from all this infrastructure, built and maintained by our College.  The instance described above sits within this environment.  Once the server is up, we don't really do much more-- the data is automatically backed up, and the system just runs.

If you look at the talk from Imperial College (see above), one of the challenges they have highlighted is getting people to understand a different way of working with their data, and what benefits it can provide them.  We've found that having someone who can show people what OMERO does is very helpful.  The other critical thing to remember is that OMERO, out of the box, is infrastructure.  It provides some capabilities, but is designed for your own extensions, via Matlab, ImageJ, Java, Python, C++ and so on.  Our own usage took off when one of the people in my lab who knew some Matlab started building tools that automated alot of the segmentation and measurement that we need to do with our data.  These tools are custom-- they solve our problems.  They took a month or so to build, and have been running ever since.  OMERO allows these Matlab scripts to work on large sets of images, and dump output in convenient ways.  That was perhaps the most important thing we have done to drive use of OMERO in our own labs.  Now it is going viral, forcing us to solve the storage issues (described above). 

Finally, in general, getting into this business (regardless of what you decide to use) means having someone who has the knowledge to run these type of services.  At the very least, getting them set up, configured,  and running takes some know-how.  There's alot to be said for bringing someone to your site.  However, you can also purchase time from consultants to do this.  We have a start-up company who can provide this type of service-- see http://glencoesoftware.com.  Or you can purchase an installed, integrated, running Columbus® system from PerkinElmer.  DECLARATION OF CONFLICT-- I founded Glencoe Software, and Glencoe Software licenses OMERO to PerkinElmer for Columbus®.  Nearly all Glencoe Software's development work is contributed to the open source OMERO codebase, for use by the community, under the GPL license.

How long to install OMERO?  With a vanilla install of Ubuntu 10.04, I followed the instructions at 

http://www.openmicroscopy.org/community/viewtopic.php?f=5&t=665

and was up and running in about 25 mins.  The bulk of the work is getting the system integrated into your own environment, especially with the tools that your scientists  use for image analysis and processing.

Hope that helps.  

Cheers,

Jason

On 22 Jun 2011, at 18:48, Brian Westra wrote:

> Hi everyone, 
> 
> I'm new to OMERO and this list. We're planning on doing a pilot case with one of our research labs and then evaluating things from there.
> 
> A couple of newbie questions, and my apologies up front if these have been covered in previous posts.
> 
> 1) How many of you are at an academic institution and in "production mode" with OMERO?
> 
> 2) Have you installed multiple instances of OMERO, or in some other way scaled up your installation to manage image data across multiple centers/institutes and research labs? Whichever approach you've taken, are there any lessons you've learned that you can pass along?
> 
> 3) We're very thin in terms of technical support that can be devoted to OMERO set up and maintenance. Approximately how many hours/week (or month or year) does your installation require for sys admin and/or other maintenance activities (feel free to delineate the types of activities, such as account setup, client and server software updates, problem tracking/help tickets, etc.)
> 
> Thanks,
> Brian
> 
> -- 
> Brian Westra  |  Lorry I. Lokey Science Data Services Librarian
> University of Oregon Libraries
> 541-346-2654  |  bwestra at uoregon.edu
> _______________________________________________
> ome-users mailing list
> ome-users at lists.openmicroscopy.org.uk
> http://lists.openmicroscopy.org.uk/mailman/listinfo/ome-users

**************************
Wellcome Trust Centre for Gene Regulation & Expression
College of Life Sciences
MSI/WTB/JBC Complex
University of Dundee
Dow Street
Dundee  DD1 5EH
United Kingdom

phone (01382) 385819
Intl phone:  44 1382 385819 
FAX   (01382) 388072 
email: jason at lifesci.dundee.ac.uk

Lab Page: http://gre.lifesci.dundee.ac.uk/staff/jason_swedlow.html
Open Microscopy Environment: http://openmicroscopy.org
**************************

The University of Dundee is a Scottish Registered Charity, No. SC015096.

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openmicroscopy.org.uk/pipermail/ome-users/attachments/20110626/a48f1e1b/attachment.html>