HCLSIG BioRDF Subgroup/Meetings/2006-11-27 Conference Call
From W3C Wiki
- Date of Call: Monday November 27, 2006
- Time of Call: 11:00am Eastern Time
- Dial-In #: +1.617.761.6200 (Cambridge, MA)
- Participant Access Code: 246733 ("BIORDF")
- IRC Channel: irc.w3.org port 6665 channel #BioRDF (see W3C IRC page for details, or see Web IRC)
- Duration: ~1 hour
- Convener: Susie Stephens
- Scribe: Susie Stephens
- Housekeeping: Papers & Data.
- Select conferences for presenting work.
- Select scientific queries.
Attendees: Scott Marshall, Kerstin Forsberg, Alan Ruttenberg, Vipal Kashyap, Bill Dyer, Kei Cheung, Joanne Luciano, Oliver Bodenreider, Bill Bug, Helen Chen, John Barkley.
Regrets: Marja Koivunen
SS reminds people to post their RDF data onto the BioRDF Wiki. She also reminded people to post any new articles to the document section, and to keep task force information up to date.
The group discussed the best conferences to present the demo at, based upon the conference information that Bill Bug had posted onto the HCLS wiki. http://esw.w3.org/topic/HCLSIG_BioRDF_Subgroup/Conference_Possibilities
The group decided that it would be good to show a demo at ISMB, but given that the deadline for proposals isn’t until May 2007, that it wouldn’t provide much motivation to getting the work done.
BB? said that we should collaborate with the BioOntology SIG at ISMB.
JL to send a mail to Robert Stevens to see if he would be interested in collaborating.
AR said that we should all think of possible ideas for collaborating.
BB said that a possibility could be a better link between RDF and OWL.
JL said that she believes that the theme for the BioOntologies SIG this year is eScience.
JL said that data integration has been around for a while as an issue, but that if data is provided in RDF that the integration wouldn’t be such an issue, so a focus could be on making data available.
AR suggested that it may be interesting to provide the SIG with real experiences gain with data integration using semantic web technologies. It’s not always so easy when you try it out, for example, data integration at the F2F didn’t go so smoothly. Just making data available is only half the story, we need to use the data sets together.
SS said that the deadlines for WWW conference had already passed. However, she’s spoken to Ivan Herman about the WWW Conference, and he said that he thinks it should be possible to secure a slot for us within the W3C track.
VK mentioned that Kei has submitted a workshop proposal to the WWW Conference, so if that’s accepted, it could include the demo. Co-authors include SS, VK, JL, etc.
SS said that it’s important that there is a reference to the demo in the workshop description, if we are going to show it within a session.
KC mentioned that he had not yet heard whether the submission was successful.
SS said that the deadline for the AMIA medinfo is Dec. 4, which means that we don’t have enough time to submit a proposal.
OB said that he would be presenting at the Medinfo meeting with Satya on the Entrez gene work, and that he plans to reference BioRDF. J
SS said that no data is available yet for deadlines for the AMIA Spring Congress, so this is something that we should keep an eye on.
OB said that the AMIA Annual Symposium may be a good meeting to aim for. The deadline for submission is March 15, 2007.
SM mentioned that a faceted browser had won the demo challenge at ISWC, and was curious to know if anyone had seen a demo. http://e-culture.multimedian.nl/demo/facet
It appears that no one had seen the demo.
JL mentioned that Siderean offers similar technology, and that she’d worked with SS on a paper that used their software for life sciences.
SS asked if it was important whether the software is open source.
AR believes it is important that the software isn’t only commercially available, as it’s important that others can download the demo.
JL agreed that we don’t want to use commercial software.
JL will see if Siderean will make their software available as a demo.
BB does the software use OWL as well as RDF.
SS said that Siderean just uses RDF.
SM said that the demo software is also just based on RDF. Although it could be extended to support OWL.
BB said that that could be a problem.
SM to see if the group in The Netherlands would be interested in giving us a demo.
BB said that he’d sent a URL to the list of the use cases that BIRN is working on, and that he will re-send the mail. Most of the use cases focus on schizophrenia. He feels that the SWAN team would be the best group for providing uses cases for Alzheimer’s. BB will identify some of the best use cases, and will go over them during the next BioRDF call. There are use cases for Parkinson’s disease, but they use a mouse model. Huntington’s isn’t a major focus for the use cases, but there is a huge Huntington’s focused group.
AR said that he felt it was important to look for questions that could be answerable with the data that we have, rather than looking for a particular disease.
BB said he’d look for data sets in BIRN that could be used for a demo.
SM said that it’d be nice if the query resulted in some interesting visualization.
BB said that you can take data sets of any kind, e.g. images, microarrays and location. You could then show data integration across the 3 mapping systems. It could be possible to add new expression data and to show it in a spatial context.
BB said that there are interesting stereological studies, i.e. using statistics as a means of sampling in space. This is big in digital anatomy. This approach could be used with microarray data and an atlas system for showing all genes that are expressed to a certain level with the substantia nigra; or for measuring differences in sizes of brain regions.
BD gave an overview of the clinical trial practice. BD will post a workflow of the process on the wiki. Data is collected at the bedside in a hospital for example. The data may be collected in paper format. The data goes into a sponsor’s database where it is cleaned. If a CRO is used, then there is an extra level of representation of the data. Once the data sets are collected and cleaned, they are entered into a data management system. This is where coding is done with MEDRA and WHODrug. Adverse event data goes into yet another database, and is handled by a safety group, that is separate from the data management group. Data is typically extracted as views, and snapshots that are typically taken after batch validation. Statistics may look at the snapshots. At the end of the process, data is put into SAS.
VK asked if LOINC or SNOMED get used in the process.
BD said that many can use LOINC and SNOMED. Typically people use MEDRA and WHOdrug. There are ongoing discussions with other institutes about using SNOMED. The thrust is towards CDISC and HL7. They are looking at other terminologies.
VK asked how HTB, Oracle Clinical and the RDF Data Model are integrated.
BD said that there are ongoing discussions with research centers. HTB is a development platform, and Oracle development is working to build applications as web services that can utilize the architecture.
BD said that the clinical trial and clinical practice worlds are very different. One is results focused the other is patient focused. They use different vocabularies.
BB said that SAS is a huge clinical component at BIRN. Deidentifying data is also a large focus. This isn’t linked to clinical trials. It’s helpful to consider this when designing the bench to bedside demo. We should make sure that we tap into existing work, e.g. NCI.
VK could be useful in the ability to apply different views to the same data. Depending upon what people are interested in, people could view the clinical trial component or the clinical component.
HC said that we should incorporate guidelines into the demo. But how big is the scope.
VK responded that it would be good to get one guideline into the picture. He won’t design a huge ontology.
HC said that it is important to represent clinical workflow and clinical records.
- JL to send mail to Robert Stevens to see if he would be interested in collaborating.
- JL will see if Siderean will make their software available as a demo.
- SM to see if the group in The Netherlands would be interested in giving us a demo.
- BB to send out the URL for the use cases again.
- BB will identify some of the best use cases, and will go over them during the next BioRDF call.
- BD will post a workflow of the clinical trial process on the wiki.