Important legal notice
Contact   |   Search   
RTD info logoMagazine on European Research Special Issue - April 2005   
 In Brief

Download pdf de en fr

Title  En route to e-biology

The phenomenal increase in computer memories has already permitted research in fields such as genome sequencing, DNA and protein sequence comparison using biocomputing, and the compilation of brain imagery databanks. We could now be entering a new phase with the development of remote calculation that permits a sharing of computing power.

The aim of the EU-funded DataGrid project is to develop a grid of remote computers to offer scientists an unprecedented calculation and data storage resource. The project is being conducted by CERN with 20 other scientific and industrial partners.   © CERN
The aim of the EU-funded DataGrid project is to develop a grid of remote computers to offer scientists an unprecedented calculation and data storage resource. The project is being conducted by CERN with 20 other scientific and industrial partners.
A few millennia ago, the invention of writing enabled us to record some of our knowledge on papyrus, parchment or clay tablets. A few centuries ago, the invention of printing rendered this recorded memory stable and, thus, more reliable. More recently, computers have taken up the task, but with one major difference: the computer does not just ease the burden on our memory, it also enables researchers to ask new questions and to embark on previously unimaginable projects. 

The saga of the human genome
The best example is no doubt human genome sequencing. This would not have been possible without the powerful machines that electronically assembled and sequenced the fragments of our genome in dozens of laboratories across the world. This initial success speeded up the development of a new discipline: biocomputing. This technology makes it possible to benefit from the ever-improving performance of computers to compare DNA sequences (between individuals or between species), to identify the regions that code proteins and also those infinitely more numerous regions that do not.

Biocomputing has also revolutionised a long-standing problem of biochemistry, namely how to uncover the spatial form of a molecule on the basis of a protein’s amino acid sequence. Until the 1990s, researchers crystallised the protein and observed it by means of X-ray diffraction. This long and painstaking task was almost impossible in the case of insoluble proteins. Using biocomputing, we can now predict – to a certain degree – the three-dimensional organisation of a protein and subsequently formulate hypotheses regarding its function. 

Imaging and the brain
The increase in calculating power has also sparked some spectacular developments in biological research. Neuroimaging provides incredibly precise images of the brain, its structure and its in vivo functioning. Its flagship technology and functional magnetic resonance imaging (fMRI) make it possible to visualise which regions are activated in response to cognitive stimulation, such as making a mental calculation, listening to music, or thinking of holidays. A single fMRI experiment gathers 2 gigabytes of data.  

For several years now, the Organisation for Human Brain Mapping has been thinking about ways of sharing this wealth of data. Some dream of a human brain programme, similar to the one that permitted the decrypting of the genome, that would seek to provide a full and systematic description of the brain.

“Human genome sequencing was a huge task, but quite simple at the conceptual level,” points out Richard Frackowiak of the Wellcome Department for Cognitive Neurology in London (UK). “But before building a common neuroimaging base, we must carry out fundamental work to arrive at a standardised methodology both for obtaining results and their processing and storage."

From DataGrid to Egee
Implementing such a plan also poses the technical problem of how to obtain the computational power needed for managing such enormous data volumes. Computer scientists came up with the concept of grid computing which networks computers so that they can work collectively on the same project. This also provides an economic solution to increasing their storage capacity. CERN (Geneva), supported by the EU, was a pioneer of this approach in launching the DataGrid pilot programme on which 21 institutes from 11 countries(1) co-operated between 2000 and 2004.

"DataGrid provided European scientists with the first convincing demonstration, on a large scale, of an operating grid,” points out project coordinator Fabrizo Gagliardi. This success, which benefited physics, as well as biology and medicine, prompted the partners to move up a gear. The result was Enabling Grids for e-science in Europe (Egee), a network of 70 European laboratories designed to provide computing power equivalent to that of 100 000 PCs available 24 hours a day, seven days a week.

"Egee will permit reliable and regular access to this technology for all European scientists, as well as for the R&D industrial sector. Just like the World Wide Web, initially developed at Cern to cover its specific needs, the precise impact on society of this emerging technology of grid computing is difficult to forecast. But it is likely to be huge,” concludes   Fabrizio Gagliardi, who is Egee’s current project leader.

(1) Germany, Denmark, Spain, Finland, France, Hungary, Italy, the Netherlands, the United Kingdom, Sweden and the Czech Republic.

Printable version

Features 1 2
  Computers, memory and thought
  En route to e-biology

  • Data Grid
  • Enabling Grids for e-science in Europe

    Features 1 2