Computer science should be placed on the same footing as natural sciences to to make progress on understanding complex phenomena according to a new report.
2020 Science was written by 34 scientists and led by Microsoft's research branch in Cambridge, England. Microsoft has announced it will contribute 2.5 million for research in specific areas outlined by the report.
The 82-page report is a forecast of how computers will become an integral part of intelligently sorting and digesting reams of scientific data. Areas of interests include "prediction machines," algorithms that let computers make predictions based on complex data and "codification," the ability to create software programs based on biological processes.
"The essence of our findings I think is the ability to tackle these challenges is about to be transformed by entirely new kinds of tools and approaches in computing and computer sciences," said Stephen Emmott, director of Microsoft's European scientific research programmes and chairman of the 2020 Science Group. "Scientific revolutions don't occur very often."
Microsoft's hope in backing the research is to tap the software opportunities that might exist in new businesses created by new kinds of science, said Andrew Herbert, managing director of the Microsoft Research Centre.
Scientists are simply becoming swamped with data. Over the last 12 months, more data has been collected than since the beginning of science, said Alexander Szalay, a professor of astronomy at Johns Hopkins University.
"There is this explosion of data and scientists, like it or not, they have to cope with it," Szalay said.
The onslaught of data has to be dealt with by the scientific process, but it is increasingly difficult as gigabytes of data has grown into terabytes, Szalay said. The amount of data will eventually exceed the raw computing power capable of absorbing it, so new tools and algorithms are needed to analyse it, he said.
Education will also play a key role in how computer science is integrated into the natural sciences, said Andrew Parker, a professor of high-energy physics and director of Cambridge University's eScience Center. Parker said his doctorate students lack training in data handling and analysis, inference and statistics.
"They need computational science courses which are relevant to analysing large data collections," Parker said.
Neil Ferguson, professor of mathematical biology at Imperial College, in London, said computer science has the potential to create a "global epidemic simulator" that would be able to recognise disease patterns derived from various data sources - such as public-health systems, news reports, and the movements of animals and people - to give an early warning of epidemics.
"We might have this narrow window where we could intervene to stop disease," Ferguson said.