Progress Report - May 2006
The Integrative Biology project is developing tools and services to help life scientists build and manage the computer simulations they need to advance their research.
By providing easy-to-use access to large scale compute facilities, the project has enabled researchers to run their existing simulations much faster, to extend these simulations to implement more detailed models and to generate significantly more data, thereby transcending the limitations of their own desktop or group facilities. Project partners have access to the UK National Grid Services which include more than 2000 processors and over 36TB of data storage capacity. In addition, some partners have access to the HPCx facility in the UK which is a tightly coupled high performance supercomputer composed of 1536 IBM P5-575 high end compute nodes. This facility is designed for those whose models run on multiple processors and have been optimised for capability computing.
The project has recently released its first integrated prototype software environment for life sciences research. The portal to this environment has been developed in collaboration with the JISC-funded Integrative Biology Virtual Research Environment project. This portal enables users to submit jobs to a variety of compute facilities on the Grid, to monitor their progress and to use the project's secure data repository based on the Storage Resource Broker (SRB) to automatically store input and output data files. SRB can also be used to store simulation codes so they may be downloaded and installed onto target machines.
Prototype software environment for life sciences research
Researchers are able to manage their work using a metadata system which understands the concept of computational experiments and builds the links needed to connect together all the relevant information about an experiment. Provenance metadata is automatically captured about a simulation run and stored along with the results for later use. This greatly simplifies the problem of managing the huge volume and complexity of information involved in life science research.
Researchers are able to manage their work using a metadata system interface
The project has also worked with some of the partners to adapt existing visualisation tools so they can be used to efficiently analyse the very large 3D datasets created in these simulations. Visualisation developments include efficient cutting-plane algorithms, iso-surface generation and movie generation from animation of 3D datasets.
Examples of the experimental output from the demonstrator system.
These new developments by the project are now offering individual researchers secure access to large scale computing, data and visualisation facilities and are increasing their opportunities for working more collaboratively with colleagues through being able to share information.
If you would like to learn more about the Integrative Biology project, please contact pm@integrativebiology.ac.uk