Blogs

Ontology Builder Tool for i2b2

Hi All

Here at Brisskit, we have been working on prototyping an ontology builder tool for i2b2.

We have developed an early prototype, its built using Java, REST, JQuery and bootstrap.

A demonstration is available at this url http://bru2.brisskit.le.ac.uk:8080/OntologyBuilder/create.jsp

This youtube video details how to use it :

http://www.youtube.com/watch?v=Y_mg0KoxrLM

Its a first phase release, enhancements planned include the ability to load and edit existing ontologies, assign nodes to integer and text metadata, create batch excel uploads.

We will be shortly releasing the code as open source.

Many thanks

Saj Issa
brisskit@le.ac.uk

Tags: 

September 2013 News update on BRISSKit

Hi everyone,

I wanted to provide a short update on the continuing and very positive progress in the BRISSKit project including a new investment from the Cancer Theme BioBank at the University of Leicester.

Some of you will be aware we have moved the project (and some of the personnel including myself and Olly Butters!) into Prof Paul Burton’s D2K Data to Knowledge research group based in the Department of Health Sciences. The D2K group has been based at the University of Leicester but is currently moving to be based in the School of Social Medicine at the University of Bristol. However, for the foreseeable future a significant part of the group (including myself) will remain based in Leicester and continuing to work closely with colleagues in the Joint Biomedical Research Unit (BRU) Informatics Data to Knowledge for Practice theme, comprising researchers from the Department of Genetics as well as the Cardiovascular, Respiratory and Lifestyle BRUs.

In Bristol, BRISSKit components will be implemented where appropriate in support of the major Avon Longitudinal Study of Parents and Children (ALSPAC) birth cohort studies hosted at Bristol, and in partnership with other South Western and Welsh university partners including the Swansea based E-Health Informatics Research Centre (eHIRC) centre (one of 4 health informatics institutes comprising the new UK-wide Farr Institute). We have proposed to build on this work through a joint bid to the MRC Medical Bioinformatics Call: building capability, capacity and infrastructure. The bid is led by the University of Bristol in partnership with Leicester, Cardiff, Exeter and Swansea. If funded, BRISSKit will lead the implementation of the data integration work package in 2014/15 and beyond.

I want to highlight in particular a new commitment by the Cancer Theme BioBank at Leicester. The group currently has associate registration with the UK BioBank consortium organised through Cancer Research UK. They are in the final stages of a bid to become the East Midlands hub for the UK BioBank. The College of Medicine at Leicester has agreed to underwrite the commitment to BRISSKit for tissue sampling for a minimum of two years in support of local and later regional needs for this group.

We are awaiting news of a proposal to Jisc to provide funding to implement BRISSKit as a Jisc endorsed national service with appropriate professional support and a legally defined, consortium type organisation that we hope many of you will be interested in joining.

So do contact us via brisskit@le.ac.uk with any queries, suggestions or demo requests - both in person or remote. To this end I will be inviting previous participants at our workshops to new JISCMail BRISSKit-announce and BRISSKit-discuss mailing lists as we build our community. We will of course also be engaging wherever possible with related initiatives including eTriks and TranSMART.

Many thanks to everyone for their continuing support,

Jonathan Tedds, PI BRISSKit

3000 Subversion commits

We reached 3000 subversion commits yesterday, and to celebrate I made another gource video. I know when we hit 2000 commits a few months ago I said I'd next do it at 5000, but I couldn't wait that long! Anyway, enjoy.

Tags: 

BRISSKIT DEMO

Youtube video demonstrating the BRISSkit stack of open source software.

European i2b2 Academic User Meeting

Saj Issa, Nick Holden and Richard Bramley
attended the very first European i2b2 Academic
User Meeting in Nuremberg, Germany on 25th March 2013.

A highly successful event organised by the Medical Informatics Group, Erlangen. Notable attendees included Uli Sax & the team from Universities of Erlangen & Goettingen and Shawn Murphy of i2b2. Gained insight into European based i2b2 developments.

Significant i2b2 plugin development has been done by the Fondazione Salvatore Maugeri hospital in Pavia, Italy. They have numerous i2b2 plugins which they are releasing as open source. Sebastian Mate of Universities of Erlangen & Goettingen has continued developing his configurable i2b2 installation routines An interesting demo was given by the University of Paris of their clinical and translational tool for researchers called Roogle and its proposed integration with i2b2.

Much networking was done and valuable contacts were gained, a description of the Brisskit components was given and a live demonstration of the Brisskit API upload to i2b2 was performed.

Shawn Murphy, one of the founders of i2b2, acknowledged and praised our strategy, he also recognized the need of a common i2b2 API.
In other news the latest version of i2b2 v1.7 will be released in April 13 for evaluation. Overall the European i2b2 community is growing and there is willingness to work together.

Tags: 
Malcolm Newbury's picture

Stratified Medicine was Cool

The Leicester Space Ideas Hub excelled itself yesterday, with a packed i day conference , called Stratified Medicine

http://www2.le.ac.uk/projects/space-ideas-hub/events/upcoming-events/str...

at the Welcome Trust building in Leicester.

Brisskit Project attended in strength to let researchers know about Brisskit's end-end research data management capability operated at the UHL Glenfield site.

Speakers and delegates included key UK health innovation leaders, researchers and hopeful SMEs.

Great buzz, good food and plenty of meetings to plan going forward.

2000 Subversion commits!

This week BRISSKit hit a bit of milestone - 2000 commits to our Subversion repository. If you don't know what a Subversion repository is then just think of it as a central storage facility where we keep all our source code. It's an integral part or our development arsenal that allows us to stay up to date with each others' code development, and gives us peace of mind that our code is stored off-site somewhere.

For those of you that are a fan of boring stats - our first commit was on 30/9/2011 by me and was rather verbosely commented as 'First import', our 2000th commit was today (by me) with a summary of 'made civi drush install more verbose'. That means we are averaging about 4.2 commits per day.

Stats are all well and good, but videos are better. This youtube video is a graphical representation of each and every commit we've made summarised in a little over three and a half minutes.

It was made with an open source piece of software called gource which reads through each commit log and adds the changes to the diagram. The key on the left hand side shows the number of files in the repository with a given file extension, so gives an indication of what kind of files we work on most.

I find these videos strangely hypnotic I have to admit. Maybe I'll post another video when we hit 5000 commits!

Malcolm Newbury's picture

Brisskit team rock up to Wellcome Trust Hack

Got off to a slow start, but Jeff Lusted, Saj Issa, Andrew Roberts and myself have finally got our teeth into a useful problem here at the Wellcome Trust sponsored Rewired State #openscience hack twitter #wthack.

We are looking at uploading the MIMIC II ICU data into Brisskit's i2b2 instance, to make the data more logical and accessible to researchers.

MIMIC is quite a complex relational database - designed for data collection but not for analysis. I2b2 is of course designed for analysis, so the migration of all of this data will be impossible in 2 days, but if we can achieve migration of a core portion of it, involving the MIMIC II UCL peeps, then this will be a significant achievement by the team.

In addition, we are looking at installing and loading the postgres port of i2b2 from transmart - now that would be extremely cool!

Malcolm

BRISSKit overview and status brochure

We have a new BRISSKit brochure (5MB PDF) describing the project and example implementations as well as setting the context of the use of mature open source applications in biomedical research as a shared service or installed locally.

Following on from the successful health research hack event in October 2012 we hope to keep up the momentum in this fast changing area - that means we will continue to flag up new projects we are involved in with partners piloting BRISSKit, but we also want to encourage everyone to contribute likewise - doesn't have to be BRISSKit specific!

Direct email enquiries about piloting and partnering with us to brisskit@le.ac.uk are also very welcome :)

Spot On London 2012 App Slam

Just presented the #BRISSKit biomedical research database applications http://ow.ly/fd4v6 to the Spot On London 2012 #solo12 meeting at the Wellcome Trust in the "app slam" #solo12DS slot http://www.nature.com/spoton/2012/10/spoton-london-2012-app-slam-hosted-... . Video is now online with 5 mins on BRISSKit from approx 26:30 mins in at http://www.nature.com/spoton/spoton-media/spoton-london-2012-video-tools...

The session was hosted by the Digital Science group run by Nature Publishing Group. Interestingly, other than BRISSKit which is designed to help integrate open source research database applications into biomedical research workflows, the other 5 or 6 initiatives presented nearly all focussed on research collaboration tools such as Colwiz (spin out from University of Oxford) and the hosting of so-called grey research data such as poster presentations (by FigShare and F1000). Here is a recent press release from Digital Science referring to an agreement between FigShare and publisher F1000 http://www.digital-science.com/pages/press-releases#figshareF1000 - the first publisher to integrate FigShare in this way.

David Shotton (University of Oxford) asked a good question about FigShare licensing: Mark Hahnel explained that it would be CC-BY except for datasets which would be CC0. Good news I think!

Also shown were Grokk for semantic search; the Scrazzl comparison engine for finding and evaluating products and materials for research; WikiPathways for sharing gene pathways and my favourite new one: the EmotionSense Android libraries for smartphones to assist the many behaviour management type research studies.

Pages