All About Multiple Sclerosis

More MS news articles for October 2002

Screensavers crack medical puzzle

http://news.bbc.co.uk/1/hi/sci/tech/2349247.stm

Tuesday, 22 October, 2002, 10:09 GMT 11:09 UK

The spare capacity of thousands of computers has helped scientists solve a complex problem - which could one day help them fight disease.

It is one of the first occasions so-called "distributed computing", in which each volunteer machine is given a chunk of data to compute, has led to a research paper published in a top scientific journal.

Problems suitable for "distributed computing" are those which would take years of processor time if carried out on just one, or a small group of computers.

However, if the task is divided between many thousands of computers, the time it takes to finish the job falls dramatically.

The downloaded software swings into operation when the computer has been idle for a set period.

The principle has been used for everything from the design of new drugs to the search for extraterrestrial life.

Folding conundrum

The success has been achieved by the Folding@home project, run by scientists at Stanford University in the US.

It is looking into proteins - essential chemical messengers which control many vital body functions.

Each long protein molecule is a sequence of amino acids folded into a complex, three-dimensional shape which is key to its particular role.

Protein misfolding is thought to play a role in many diseases, including CJD and Alzheimer's.

The aim of the Folding@home project was to simulate just part of this folding process, which takes just a few microseconds to happen.

A single average computer would take all day just to simulate one nanosecond of protein folding.

Willing volunteers

Folding@home was launched two years ago, and has so far recruited 200,000 PC owners.

A new recruit will download data analysis software, then be assigned particular computational tasks, sending the results back when they are completed.

A group of 30,000 computers was able to perform 32,500 folding simulations and accumulate 700 microseconds of data.

The results - predicting that a particular protein would take six microseconds to fold - tallied well with laboratory tests.

Dr Vijay Pande, from Stanford University, said: "These experiments represent a great success for distributed computing.

"Understanding how proteins fold will likely have a great impact on understanding a wide range of diseases."

The results were published in the online version of the journal Nature.

This is by no means the first success for distributed computing - it has cracked complex mathematical problems before - but it is the first to be published in a journal such as Nature.

The most famous project, the Search for Extra-Terrestrial Intelligence (SETI), which analyses radio telescope data, has also thrown up some promising "leads".
 

© 2002, BBC