Eden

Eden is about creating distributed solutions to modern bioinformatic and medical informatic problems.  The focus is not just on grids or peer-to-peer systems, but on the creation ofinnovative solutions that can be distributed among today's heterogeneous networks, both in terms of computation and of data.

Vision Statement

Algorithmics and high-performance computing are transforming research in biology and biotechnology. Today, the building of in silico models of life is a grand challenge for both computer science and biology. Deciphering the human genome has fired international science and captured the public imagination in a modern revolution. This revolution spans from the DNA sequence of a gene, to the structure of the protein for which it codes, to the activity of that protein and its function within a cell and the whole organism. This leads to the quantitative study of biological processes at a system-wide level by considering whole biological systems, instead of treating their constituent parts and processes in isolation.

Eden aims to transform the science of computational biology by providing the sophisticated tools required to solve complex biological and biomedical problems that arise naturally from the explosively increasing amounts of bioinformatic data that are becoming available.  Progress in this field demands innovative solutions to the highly data- and compute- intensive problems such data induce.  These solutions must combine advanced computing, mathematics, statistics, algorithms and data-management techniques.  The project involves computer scientists working closely with life scientists, mathematicians and medical researchers.  The synergy this generates promises immense rewards, with the potential to profoundly improve our understanding of biological systems from the molecular to the global scales, and where humans, the most powerful species ever, fit in those systems.

Short Term

At present, members of the Eden team are researching and developing a distributed BLAST (Basic Local Alignment Search Tool) algorithm, that uses distributed data in a peer-to-peer network. BLAST is one of the most commonly used bioinformatics queries and is very data intensive: caching the data is therefore crucial in order to maintain fast queries with ever-increasing amounts of data in the public databases. This distributed BLAST and the underlying architecture it uses are the test-bed for further projects to "plug in". Other projects include protein structure prediction, which is compute- and meta-data- intensive, and so will also benefit from this peer-to-peer caching system.

Long Term

In the long term, Eden will produce a suite of bioinformatics algorithms and applications that will use the distribution system currently being developed for distributed BLAST. Our aim is not just to rewrite current applications to run in parallel, but to intelligently parallelise the algorithms themselves. Bioinformatics problems often have untapped structure that will be utilised to create faster, more reliable, and scalable solutions for bioinformatics problems.