Skip to main content
Topic: databases for metadata - what systems do people use? (Read 10949 times) previous topic - next topic

databases for metadata - what systems do people use?

Hi all

I\'d like to pick up on something Susanna and Adam wrote about - databases.

We\'re at a crunch point in my lab and need to put a system in place to manage our data from info about source organism to where the NMR or MS data files are. There are commercial LIMS systems available but I\'d be interested in hearing more about what everyone else is using.

Our lab also includes proteomics and any solution we put in place would need to be compatible (at least at the interfaces) with our colleagues in genomics (who are in the process of changing their LIMS). We are looking at between 3000-10,000 samples per project and any LIMS will be managed on the department\'s cluster (unix).

isa tools seems to be something a few people have talked about. I\'m not a data base expert but I\'ve tried to have a look at some of the documentation on this. Is it an relational database that is customisable? Is this working well for people? Could we get it to recognise and/or produce bar codes? The ontology and experimental descriptors look good and certainly seems like it would work for our systems biol work.

What are other people using?

cheers
Simone

databases for metadata - what systems do people use?

Reply #1
Hi Simone,

In the Wishart group at the U of A we generally use customized systems. We\'ve tried many of the \"customizable\" solutions, but find they generally never meet the needs of our group, and the interface can be rather unintuitive. A moderate level of customization seems to be enough to get us where we want to go. That being said, I haven\'t had the opportunity to try the latest versions of things like SetupX, and I\'ve heard they are getting quite mature and easy to use.

We have being using a custom system for our metabolomics center based on Ruby on Rails. It is more like a web application, with a huge focus on usability. We are adding barcode support as it\'s quite easy to do (there are great plugins that let you just install a barcode generator with a few commands). We\'ve also managed to integrate the tool with MetaboAnalyst, a data analysis tool.

One thing I should mention is that most LIMS systems don\'t really need a cluster behind them. Unless you are doing massive amounts of computation, a generic, cheap Linux box (with good backups) or a secure (encrypted) offsite host should do the trick. Not sure if you need to pay for the cluster use, if so it could save you some money.

We are considering open-sourcing our center\'s LIMS/sample tracking system once we get the code a bit more mature. If you would be interested in this, maybe we can figure something out.

Craig

databases for metadata - what systems do people use?

Reply #2
Hi Craig

Sounds really interesting. I haven\'t heard of RoR before. When do you plan to release the open source? Do you have a controlled ontology built in as well? I\'d be interested in following up!

cheers
Simone

databases for metadata - what systems do people use?

Reply #3
Sorry for the late reply, December was a busy month  ;)
We will probably be open sourcing it in the summer, depending on how our other projects go.
Craig

 

databases for metadata - what systems do people use?

Reply #4
Dear Simone,

A study capture framework and matching metabolomics databasing module have been developed by the Netherlands Metabolomics Centre (in collaboration with other groups in the Netherlands). You can get a flavour of functionalities and test an implementation of the database framework at http://test.nmcdsp.org
For an implementation of the framework at your own lab you can download the source code from the following sites:

The study capture framework: https://trac.nbic.nl/gscf/
The metabolomics assay module: https://trac.nbic.nl/dbxp-metabolomics/

It is an open source initiative, and still development takes place (mainly on the \'looks\' of the metabolomics module and advanced querying), but upgrades will be compatible with older releases, and not influence the data already stored. The study capture framework site also refers to other modules that can be connected to the study capture framework (e.g. to store transcriptome data).

We are also implementing a Galaxy based data processing toolkit, which will include a functionality to upload data from the database framework. This is work in progress (will be extended with new tools in the near future). If you are also interested in this tool, see http://www.metabolomicscentre.nl/page/data-support-platform-0 (now only limited information, but a new website with information will be released this week).

Regards,
Margriet