From W

global links:

1. new NPD admin system (test version)
2. new NPD user system (test version)
3. Reports - weekly reports. setup and test reports, and others
4. Meetings - information on past and future meetings
5. Project plan - the original, and current updates
6. the completed dissertation
7. Original NPD - info and links
8. Web services links

---

1. The week started with a meeting with Graeme, where we discussed the prototype admin system and the possibility of pulling some data from other sources. This will need to be examined more closely, but possibilities include: gene_name-location info, protein info from the protein ID, localisation details, some of the GOID details - I will check which of these details can be easily pulled using web services.
2. This examination of web services will be important to for comparing with ogsa-dai. If alot of the other databases in the field use web services e.g. SOAP. how beneficial will ogsa-dai be? we do not need to run it over multiple databases, so there is no advantage there unless others were willing to take it up. And they will only do that if it offers advantages. It does work with web services, though I need to look into how much there is a crossover, but still, if it offers no better than other services which do not need to rely on ogsa-dai, this will not be a convincing argument. So there is a requirement to show in what way ogsa-dai is better, or not, in which case just use the other services.
3. I started the process of using ogssa-dai on Monday evening, which essentially required installing it. Whilst initial setup and evaluation included checking interoperability and dependencies, the 10-minute install procedure still managed to turn into a 10-hour install procedure. Not such a good start. The error is vague - and only results when installing ogsa-dai ontop of a tomcat container that was installed to a linux OS via the apt mechanism, even though the tomcat container worked in every other way, and gave no errors. I managed eventually to install ogsa-dai ontop of a manually installed tomcat container. I found some references online to this problem, which will be recorded for the dissertation.
4. Ogsa-dai is up, and a service to submit a query to the database is available. This offers quite powerful interaction, though it could probably be easily emulated by standard web services. The activity mechanisms in ogsa-dai allow deployers to offer methods for performing certain actions, but again this may be implementable via other means. But for now at least, the database is available via sql query via ogsa-dai.
5. I have now implemented a web service ontop of the database. This presently offers very little functionality, but was much easier to set up than the initial ogsa-dai installation, as there are protocols for web services in many languages, including PHP: hence, this web service was able to run right on the server software already in place.
6. I installed Taverna and connected to my NPD web service. That was pretty easy. I also want to try connecting to the ogsa-dai service, so I can compare how easily these things can be done.
7. Although I have contacted Marco Roos, I have not heard back from him this week. I will need to send another email I think. However, it should be possible to get an idea of how well the services are working even if he does not take part, though it would be nice to model an actual use of the data.
8. Having looked at other web services from the relevant bio-databases (listed in the Web services links section), I have successfully tested connection to some of those services, and can pull alot of data out automatically. With a bit more analysis, I should be able to specify what data could be pulled this way during the admin process. It will be necessary to have another meeting with Wendy and Graeme about this, probably in week 6.
9. The prototype admin system also needs updated with some further functionality to improve the search field in line with the user NPD. And the user NPD will need to be kept in mind when performing any changes to the database schema. For example, it will be necessary to add some ID fields to some tables, and a couple of tables are no longer required. Work will continue on this.
10. Graeme talked to the people from Gene Ontology this week, and they are interested in having a mechanism to update the NPD data into the Gene Ontology database. This will provide an additional use for web services, and could be a testable example of how the service, either via ogsa-dai or otherwise, can be used to improve data use and sharing within the field. Graeme is expecting to hear back from them soon with more details.
11. I will update the project plan to show the intended outcomes of the next few weeks.