Phone Conference Minutes
Present: Tamas, Alexey, Michael, Christian, Dorin, David, Manolis, Harry, Richard, Dmitry
Harry: After WP2, Manolis was presenting A4-A6, broadcasting on the web.
Tamas: ... on the agenda ...
Tamas: Story 3: there is a question if we want to have this as a separate story or not.
David: We need to have a distinction b/w the document and demonstration. Story 3 should go into the document, but elicitation and demonstration should focus on the clinical view, which Story 3 does not have.
If there is a story that we can show, it is how basic functionality goes into the platform. Also there is the administration use-case. There are two user communities that are providing these stories. Another thing, (which does not exist yet) is: illustrating the horizontal integration with something like a google search. That could potentially glue together the use-cases. [Manolis]: what google search? [D]: To look for whatever you want, and this would return something like patient entries.
Manolis: Use-cases versus demonstration: Agree with David about difference. Demo should be subset. Different objectives. Excercise we are doing is degrees of freedom of which component. Has to be general, etc. UC3 is orthogonal illustrates capabilities of the system.
Tamas: Disagres, the use-cases should drive the demonstration, but obviously not the demo in March.
Dorin: Let's look at practical questions. Today: what demonstrations we want to show in March? Who is working on what? Who will work with Michael to define an API to his application to the gateway? What is the mechanics? Abstraction of the use-case, define very concrete settings for building demonstrators. Michael is building an application, who is the person with the api?
Tamas: Jerome, David, Andy, etc....
David: For the EGEE demonstration Dima and Jerome were involved a lot
Dorin: Who is responsible?
David: Jerome.
Dorin: Michael and Jerome should come up with a document of the procedure detailing the connectivity. Working together means defining APIs. This is an action item. Can go into detailed specification. Other question is how many layers we have in the platform? Do we respect the structure that is on our slides? We have our HeC gateway, layers for KD, DS, IDM, etc, on top applications?
David: So far the released gateway respects D7.1. We have a couple layers, security and privacy, service access layer, providing functionality, layer for grid interface, database interface. Big layer on top where WP6 and WP8 applications come in. This layer for the moment is going to be populated.
Manolis: By the end of the year, we should have the MQL. Connecting to the ontology/query formulation engine that will be the ontology-enhanced query engine. By the end of the years, queries should be run with multiple databases.
Dorin: Michael's application requires queries. Who does this talk to? One person to be able to answer questions to
Tamas: Then it's Jerome.
Dorin: Christian is on the line... need one scenario on genomics. Genomics and imaging put together. Maybe you should start working on parallel excercises.
Tamas: We are definitely going in parallel.
Dorin: Let's have the API by next week. Other practical question: let's have a list of demonstrations/applications that we want to show in April? Should not interfere with the long-term view of the project, just about responsability. We rely on each other. Who would be the person who can compile the list of what applications we want to show in the platform working, showing everything?
David: Stories or components?
Dorin: Let's decouple. Standalone apps is much easier. Working in conjunction with the HeC gateway. Applications that are using the powers of the grid, etc.
Manolis: At the top level we have the A6 applications of three flavors. Capabilities at that level? We also have the A4 group responsible for the platform. These capabilities are orthogonal to each other. We should start from the applications and collect what we need and after that architectural features, like capability with the grid, capability with the database.
Tamas: Do you mean the inventory we discussed about earlier?
Dorin: What are we showing? Clinician comes to the review .. what happens? What are we showing? E.g. 1-2 in cardiology, 1-2 in brain, 1-2 in Rheuma. These should show the capabilities of the platform. In addition to that we plan to show this and that and that standalone applications. In addition we want to show this and that theoretical results.
Tamas: Martin should be responsible for this
Dorin: so far two actions:
Michael and Jerome with a document how the first app should be built
Martin in two weeks to have a list of what applications we want and be possible to run on the platform.
These applications can be seen from the clinical pov, but there are other applications on the technical side, like maintenance/inventory of the data, ontology focused applications. Keep in mind the user. Need to start planning and define clear responsabilities. Third point would be good for a person who makes an inventory of technologies that we have in HeC after two years of work. These technologies should be seen - we want to see the forest ... technologies that are sufficiently mature to be demonstrated live on our HeC platform, technologies demonstrated standalone, third: new concepts more theoretical on the data level, molecular, imaging, etc. Up to now we have spent 7 million, what have we got for this? Start connecting these technologies together and putting in a nice story.
Tamas: I can reluctantly volunteer.
Dorin: Martin already has one (what will be the applications that will be run on the platform). We will have two more, what would be off-line? Third: main scientific achievements that we need to present? Have to start planning, create an inventory of what we do. January is very close and we end up in January with let's start planning. We have to start very formally. First phase is to get an inventory of what we have and what each of us are thinking.
David: Maybe I can volunteer to do this inventory. We have sent out a questionnaire and we have a precise list of what is in the current release and the questionnaire.
Dorin: Using the questionnaire is a good idea, but this is not an email thing, you need to talk to people.
David: The questionnaire go into much detail even about hardware, and then we call each of the protagonists to see how this can be seen for integration.
Dorin: It would be better for you to work with Martin defining the first category, which is mostly application that work on the platform. For the other things, simplifying your work: you don't care, care as requirements but not be attached to the platform. [...] Manolis is coming with a query technology, but it needs to be matured, can't go into the platform yet.
David: At some point you will need the understanding for the integration
Dorin: We have to decouple things with three months left. Queries are part of the integration platform. But there are others, imaging of Alessandro, etc. It makes sense for you to be concerned what we show on the platform, there is little differentation b/w short-term and long-term planning. The other two points are not that important. As they are not connected to the platform, for coordination it is not that important. I hate the idea that we get one night before the presentation and ask what our project does. We have to go to a couple of rounds of brainstorming to present in the right way and another practical aspect to allow Tamas and the URS team to kind of reshape the URS in terms of what we achieved. Reverse engineering is positive.
Dorin: Can we have an initial list of technologies to be shown and scientific results that we plan to show in the March meeting. We need a champion for collecting what other applications are shown aside the platform and scientific output - aside from the clinical side!
Dorin off.
Tamas: OK, going back to the WP2 agenda. We have a nicely growing cardiology use-case, need rheuma and brain. Will ask Alessandro to kickstart the rheuma, can Christian do something about the Brain?
Christian: MLG will circulate some actions for the next months, from clinical and genetics in Gaslini. 12th of Nov - Alessandro meeting how to process, etc. This contains both processing and personalized medicine. Need to meet the next two weeks and define a kind of strategy and we will let. From Siemens we have been working with similarity search, more or less on our side the engine is usually. We select the genes by Alessandro. In gaslini there will be some more pylo astro, go deeper in the disease.
Tamas: can you try to write this up?
Tamas: What about Story 3?
David: Manolis and I look at this offline? Major concern about this story it will recall a big chunk of the previous URS. ... OTOH we won't have a very good granularity in the expression of the requirements, not a big problem, just thinking.
Tamas: ok, you do it next week
David: There is another story, story 0. We need a good description of the users community, we start building a glossary and defining a user community. That would help all the others to write down a precise vocabulary. In the end it will be very complicated and endless. Like: what do we mean by the Health-e-Child platform?
Manolis: That's right, I used for example Medical Analyst...
Tamas: No time to discuss now, deferred to next week.
Tamas: logistics: meeting in Nice confirmed, Hospital visits on the way, Please use googledocs!
Manolis: There are some unresolved questions how the use-cases equal demonstration. We might discuss this next time.
There are minutes attached to this event.
Show them.
