Integrative Biology Project Workshop 29 - 30 September 2005

This year's Integrative Biology project workshop was held at the Martin Wood Lecture Theatre in Oxford on 29th and 30th September 2005. It was attended by most of the leading participants in the IB consortium plus many of the new graduates who have recently joined the IB team. A list of delegates is available to view.

The aim of the workshop was to review progress during the past year and to solicit feedback to help guide the development programme over the coming year. The first day was devoted to a series of talks on the scientific agenda driving the project and the second day to a summary of progress to date on building the Integrative Biology Grid infrastructure and plans for the future.

David Gavaghan, the project principal investigator, welcomed delegates at the beginning of the first day and reminded everyone that the objectives of the project were to develop a computational infrastructure for in-silico modelling of the two major causes of premature death in the UK, heart failure and cancer, and to use this infrastructure to pioneer a new approach to understanding biological function. The remainder of the first day was taken up with a series of talks on the current scientific problems in heart and cancer modelling. The day 1 speakers and their slides are online.

The second day was opened by Sharon Lloyd, the project manager, who summarised current infrastructure development activities and introduced talks about the major infrastructure components and demonstrations of some of the component prototypes. There were also several break-out discussions on specific topics bringing together the relevant scientists and software developers. These provided valuable insight and feedback for the developers on user needs and priorities. The day 2 speakers and their slides are also available.

In concluding the meeting, Sharon highlighted some of the key points to emerge from the discussions:

Visualisation and interactive services: Need for visualisation resources which combine the best features of CoolGraphics and Meshalyser and add new capabilities such as isosurface generation. Computation associated with visualisation should happen where the data is to avoid large data movements. To exploit the developing capability for multi-parameter visualisation, models will need to store more variables from each time-step.
Job management: Seen as a key resource by all modellers with the current prototype providing a good basis for further development. The portal may provide an effective channel for informing users of new developments and features through a news service.
Data management: Support for developing a global data repository for holding simulation results based on a distributed SRB. With proper provenance and other metadata, this will facilitate international collaboration and provide access to the data behind publications.
Workflow: Potentially very useful and a high priority. It could replace the current use of scripts. Most workflows are linear. Data backup is essential. Ability to reproduce results through re-running workflows is a major benefit.
Steering: Relatively low interest in steering at this stage. Parameter sweeps will be useful for cancer modellers.
Virtual Research Environment: Within the VRE, workflow and visualisation are the two top priorities. Webcams to support inter-personal communication could be useful.
Code development: Support is needed for small scale as well as large scale software development. User support - Each group of users will need someone to support them.