Discovery Net: Knowledge Discovery on the Grid

Applications Overview

Fighting Natural Disasters with e-Science

Discovery Net addresses the complexities faced by scientists from all information-intensive fields where:

Discovery Net is developing a series of testbeds and demonstrators for using the technology in the areas of Life Sciences, Environmental Moonitoring and Geo-hazard Modelling.

Life Sciences: High throughput genomics and proteomics

In Life sciences, discovery net is currently being applied in
  • Real-time iCollborative Genome Annotation
  • Large-scale Integrative Functional Genomics
  • Real-time High Throughput Chemoinformatics
  • Large-scale Geneotyping Data Analysis
  • Real-time Drug Resistance Studies
  • Integrative Life Science Analysis
In addition application scientists have been using the platform for integrating and analysing data from various novel high throughput devices including Protein Folding Chips and SNP Chips using label free (LFII) technology and Protein-based fluorescent micro-array data analysis.

Our life sciences test-bed won the Most Innovative Data Intensive Application Award at Supercomputing SC2002 where we its use for the analysis of large-scale genomic data in real time using cross-continent distributed computing resources, for the annotation of malaria genomic data.

Researchers from SCBIT (Shanghai Bioinformation Institute), have successfully used Discovery Net to study the molecular evolution of the SARS virus during the SARS epidemic in China.

Environmental Monitoring: High-throughput disperesed air sensing technology

Using Discovery Net technology application scientists are building sensor grids for air pollution monitoring in urban areas. The sensor technology is based on GUSTO (Generic Ultraviolet Sensor Technology) developed at Imperial College.

The technology measures pollutants, including SO2, NO, NO2, O3 and Benzene, at ppb levels every few seconds.

Real-time Geo-hazard Modelling: High-throughput earthquake modelling through satellite imagery

Discovery Net is used to conduct remote sensing data mining for geo-hazard modelling including research studies of landslides in the 3-Gorge dam/reservoir region in China.