Accelerating
High-performance computers are opening up new ways of developing new drugs, as Beth Sharp finds out
Professor David A. Bader, executive director of High-Performance Computing, School of Computational Science and Engineering, Georgia Institute of Technology
A
s part of an award from the US National Science Foundation, we are undertaking a project to accelerate computational
discoveries at the petascale. Te project is a collaboration between three research groups – myself, Professor Jijun Tang at the University of South Carolina and Stephen Schaeffer, a Professor at Penn State – and our focus is on trying to understand the evolutionary histories of multi chromosome organisms. Te outcome will prove very useful for the pharmaceutical industry in terms of developing drug targets and understanding the evolutionary process of plants; particularly useful for determining their medicinal properties. We are developing a new algorithm and
parallel program called COGNAC (Comparing Orders of Genes using Novel Algorithms and high-performance Computers) which reconstructs phylogenetic trees using gene order data. COGNAC is a follow-on from our previous code, dubbed GRAPPA, which dealt with the evolutionary histories of
30 SCIENTIFIC COMPUTING WORLD
single chromosome organisms. Gaining an understanding of how species evolve is, as you can imagine, computationally difficult and this is where the use of HPC allows us to solve the issues efficiently and reduce the development time for pharmaceutical companies. Being able to assess human populations may also lead us to personalised medicine by offering insight into how a drug may effect one group of individuals over another. Te use of a code such as COGNAC will
enable us to differentiate people into sub groups based on their evolutionary history in order to fully know which treatments would be the most effective. Furthermore, this will help us to reduce health costs by speeding diagnosis and taking a preventative approach to disease. Te code itself is continually being developed in order to make it more useful on real data sets and for it to run faster on new HPC platforms, including multi-core Intel processors and Nvidia GPUs. By keeping up with and harnessing these capabilities, we can attempt to
solve problems that just a few years previously would have been considered intractable. Here at Georgia Tech we’re focusing on
this area with a new academic programme in computational science and engineering that addresses both computing and discipline areas such as biology and precise medicine. Tis combining of disciplines is really where we see the solution to real-world problems such as drug design and discovery, and we’re optimistic that we have the right tools and are training the best people. Te challenges lie in determining the effectiveness of these approaches and our ability to innovate with the right algorithms. Our preliminary results are good, however, and as computers become even more capable and push towards exascale, the resources will be there to make significant progress in the study and understanding of evolutionary histories.
www.scientific-computing.com
drug discovery
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44 |
Page 45 |
Page 46 |
Page 47 |
Page 48 |
Page 49 |
Page 50 |
Page 51 |
Page 52