The analysis of genomic data from this global information pool has created a new scientific discipline, bioinformatics, or using IT to study biological information.

For the first time, scientists have the means to understand the functionality of genes and their interactions at the whole genome level. The trick, as ever, is to manage the information and turn it into useful knowledge.

Last year InforSense, a won the contract to provide sophisticated data-mining and pattern recognition software to the Sanger Centre in Cambridge, the hub of the UK's contribution to the human genome project.

InforSense had foreseen that, in the emerging world of cyberspace, there was a ready market for software that could handle and evaluate information that was present not only in one database, but in a wide variety of data hubs sited across many websites.

Yike Guo, the company's senior scientist and also leader of research at Imperial College's parallel computing centre, led development of its Kensington software. "The system offers the right architecture for large-scale and distributed data analysis, which is essential for post-genome bioinformatics application," he says.

The contract with the Sanger Centre also involved the Imperial Cancer Research Foundation (ICRF) and Ludwig Institute of Cancer Research. This consortium will produce a large quantify of data relating to gene expression.

The analysis of these data will reveal vital clues of gene functions and their relations with diseases and drug effects. Such knowledge can be efficiently exploited only by an interactive discovery procedure using genomic, proteomic and other biologic information available on the internet. John Sogrous, head of the computational genomics laboratory at ICRF, and in charge of the bioinformatics infrastructure of the consortium, says: "We are now in the era of e-science - where scientific discovery is based on the global information pool distributed on the internet.

"The new generation of data analysis tools are central to e-science. Scientists need their own browser which should not only retrieve information according to your request but also analyse the information, discover knowledge from the data and retrieve more information based on those discoveries. Kensington is certainly moving in this direction."

Dr Guo says: "In scientific data there are many associations or correlations that are invaluable and highly sophisticated, but difficult to spot my manual means. For example, which gene regulates the proteins involved in cancer."

Kensington has been designed to operate beyond current techniques of data mining - "since most can only analyse static data warehoused for some fixed business logic," according to Dr Guo.

Its main task, therefore, becomes one of complex communication, organisation, analysis and visualisation of the ever increasing amounts of data becoming available, often distributed around the internet or on company intranets.

InforSense obtained $1m in financing in its first financial year, and has already sold 7 per cent of the company last year for around $1m from private investors. The company is now looking for second stage funding.

InforSense is now developing a variant of Kensington to meet the needs of e-pharmatics, providing an enterprise-wide bioinformatics system to provide systematic drug discovery based on genomic and proteomic information.

Gene Logic, a US bioinformatics company, has purchased the system to provide a core for the analysis of genomic information.

At a second biotech company, Massachusetts-based Curis, Dan Passeri, vice president of corporate development, says:"The complexities of analysis for deciphering meaningful information from genomics requires pattern recognition and interactive data-mining capabilities.

"We believe InforSense represents the next generation of analytical software for helping to discover the promises of the genomic revolution."