Computing Sciences on the Norwich Research Park
Scientific research is increasingly dependent on information technology and the Norwich Research Park has key strengths in this area. Computational techniques are used to obtain and store data, to analyse and interpret this information, and to extract new knowledge from it. Many of the NRP member organisations have extensive capabilities in computing sciences, working closely with the other strands of life science research on the NRP, as well as reaching more widely into other areas such as climate systems, speech and colour recognition and virtual reality.
A major focal point is the Computational Biology Laboratory in the School of Computing Sciences, which is providing excellent opportunities for local, national and international collaborations between biologists and computational scientists. This facility is electronically linked to a £1.5 million state of the art Henry Wellcome Laboratory for Cell Imaging. This facility enables scientists to analyse, in real time and three dimensions, cellular mechanisms responsible for conditions such as cancer, arthritis, deafness and cataracts (See HEALTH). The connection between the bioimaging suite and the bioinformatics centre enables high throughput data analysis.
Quick Links
Computational Biology
Data Mining
Computer modelling and risk analysis
Computers in vision, hearing and speech
Computational Biology
Data Mining
Computer modelling and risk analysis
Computers in vision, hearing and speech
Recent advances in technologies for generating biological data have posed new challenges for managing the wealth of information that is produced. Genome sequencing projects, for example, spawned a branch of computational biology which deals with the science of analysing gene sequence and expression data and presenting it visually in ways that help scientists understand it. Even individual experiments now yield vast amounts of data and bioinformaticians on the NRP are developing new computing tools for their analysis. The new tools being developed enable researchers to compare genetic information between different species, particularly crops, to help understand more about biological diversity and evolutionary relationships.
There is strength on the NRP in modelling the regulatory networks that organisms use to control the expression of their genes. These networks must allow expression of genes that are required for particular tissue types and stage of development. The models must also enable subtle responses of the organism to changing conditions. Ultimately the aim is develop computer programs that can describe all aspects of growth and development including the shape and size, metabolism and physiology. Such programs will give scientists a powerful tool to make predictions and suggest novel ways to tackle problems.
These approaches extend to managing information about the protein and chemical profiles of cells from different organisms generated by the proteomics and metabolomics labs. In particular, methods for the analysis of complex datasets generated from the analysis of metabolites are being developed.
The genome programmes are yielding sequences for many millions of genes, but the structures and functions for most of them can only be guessed at. Novel computational methods are being developed on the NRP that can more accurately and rapidly predict 3-D protein structures and from those structures inform scientist of the proteins' functions. These predictions can be tested in collaboration with scientists collecting structural information from X-Ray and NMR analysis who also depend on developments in computation to convert instrumentation readings into three dimensional images.
More on computational biology
Data mining
Data mining permits significant information to be extracted from large data sets, and has applications in the life sciences, and social sciences such as in insurance companies, patient records in hospitals and fraud detection with customer profile data. Such information is usually stored in a database, which needs to yield meaningful results when asked a question by the user. Finding new ways of mining this data is a large industry, in which the NRP has considerable expertise.
More on data mining
Computer modelling and risk analysis
Computing technology is being put to powerful use on the NRP to help predict hazard and risk. For example, computer-based models, known as Bayesian Belief Networks have been developed to help the food industry assess food safety issues. These models enable multiple calculations of risks, costs and benefits to be carried out in a single framework, enabling more rapid, cost-effective decision-making. The ComBase is a powerful tool aimed at providing the food industry with information about microbiological safety. Compiled by a mathematician, the database holds information about the growth of different harmful bacteria in a range of foods, for example, dairy products, vegetables and meat under a range of environmental conditions. By simply typing in the conditions specific to a particular industrial process or facility, a company can rapidly access information about the likelihood of bacterial growth in their product. This information gives a general guide for further analysis by the in-house research departments within a commercial organisation (See FOOD).
Computers in vision, hearing and speech
Colour perception and reproduction is an area of great interest to researchers on the NRP. Industrial collaborations are focussing on developing technologies to improve colour scanning technologies, and ways of determining intrinsic colours of objects in order to reproduce them accurately. Image capture and reproduction in digital photography is also a fast moving field in which the NRP has a strong expertise; by developing mathematical tools to "correct" digital images so that they appear the same as the real object does to our eye.
More on colour perception
Speech recognition software has been around for some time but there is a lot of room for improvement. This is particularly in situations where speech is spontaneous and disfluent rather than planned and fluent and there may be problems of background noise or degraded sound e.g. on a mobile phone. New computing methods to deal with these problems are being addressed.
TESSA is a system developed to help the deaf or hard of hearing during Post Office transactions, and enables cashiers to communicate more easily with hearing impaired customers. It combines speech recognition technology with virtual human animation in a system where the cashier's speech is converted into British Sign Language and signed by the computer-generated human for the customer. Working closely with a number of collaborators, including the Royal National Institute for Deaf People, the researchers have developed software to capture the movements of a real person using sign language. These movements are then stored and used to animate the virtual human when required.
Back to top
Last Updated: 09/04/2010 14:59