Systems analysis called key to unlocking bio data
By Chappell Brown, EE Times
October 20, 2003 (11:49 a.m. EST)
URL: eetimes.com
Hancock, N.H. - Biomedical researchers are eagerly anticipating a new era of "biosilico" experimentation, where slow and painstaking biological experiments will be replaced by computer-based data mining and simulation.
The prospect of decoding the link between DNA and widespread plagues such as cancer and heart disease is propelling a new segment of industry dedicated to lab-on-a-chip hardware and bioinformatics software. But the whole enterprise might founder if it proceeds on the basis of past experimental paradigms, said Leroy Hood, president of the Institute for Systems Biology (Seattle), who is proposing a fusion of systems analysis and biological experiment as a means of coping with incipient information overload.
Hood believes information mining and computer simulation by themselves will never be able to decode the complexity of biological systems. Instead, biological researchers must set up models based on systems analysis and then design experiments based on them, he said. In an iterative process, the experimental results are compared with model predictions and the computer's model is modified accordingly.
Hood will be a keynote speaker at Biosilico 2003 at Stanford University this week, billed as a "strategic summit on silico biology." Biological experiments typically take place "in vivo"-in a living system-or "in vitro," in lab apparatus. The term "in silico" has been added to the language to indicate a fusion of pharmacological, biological and computer technologies into a new experimental environment.
The reduction of DNA lab analysis to the chip level, soon to be followed by chip-scale protein-analysis systems, has generated an information explosion for biomedical researchers. The scramble is on to standardize biological information and build data-mining "bioinformatics" systems that will allow the research community to truly benefit from the flood of new information.
Molecular biology, decoding the genome, the study of proteins-all have produced fundamental breakthroughs over the past few decades, but they represent only a start in understanding the entire systems of biological organisms.
"Biology stands at this curious inflection point, where it is being driven by new technologies-DNA sequencing, multiparameter analysis, molecular recognition and new mathematical tools for analyzing data," Hood said. "But it is also going to be driven by powerful new strategies such as systems biology and predictive preventive personalized medicine." In his view, the trend is to move from the individual analysis of DNA, RNA mechanisms and proteins toward a mapping of complex interactions between cell components or higher-level functions in a bid to understand emergent behavior.
In a few years, Hood expects to see technology that can sequence the genome of an individual for less than $1,000. The next step will be a similar capability for analyzing proteins, which will provide critical information on how far down a path to a genetically programmed disease condition an individual has progressed. That will make preventive measures highly effective since the precise time to intervene will be known, and systems analysis will be able to gauge the proper therapies, eliminating side effects. That capability will personalize medicine to a degree never achieved in the past.
Systems analysis will be critical in tracking therapies and also in the design of drugs and proteins used to alter the effects of genetic deficiencies. This is a new paradigm for biological research in that whole systems, rather than specific components such as cells or organs, will be tracked to steer therapies in the right direction.
The structure of DNA was discovered in the 1950s, just as the first digital computers were being built.
"DNA represents the digital component of living systems, but there is much more," Hood explained. "DNA is modified over the immense time scale of evolution, organs develop over a period of decades and the chemical responses of the body can take place in milliseconds. In addition, there is a complex hierarchy of systems from DNA to cells to organs and on to living individuals."
Hood has played a strategic role in creating the technology that resulted in the recent triumph of sequencing the human genome. In 1985 while at Caltech, he developed the first automated DNA sequencer using fluorescent dyes, lasers and computer- controlled readout. The technology accelerated the reading of DNA by a factor of 3,000 and made the Human Genome Project a reality. Hood went on to found the department of molecular biotechnology at the University of Washington and recently left academia to found the Institute for Systems Biology.
Working with a group at MIT's Whitehead Institute, the new institute has developed an open-source software package called Cytoscape, which takes large databases of protein-protein, protein-DNA and genetic interactions and maps them into an interconnected-graph representation. Genes, proteins and molecules are represented as nodes and their various interactions by arcs connecting the nodes. The basic representational system can be extended by adding computational components to analyze activity.
Researchers at both institutes have used Cytoscape to identify gene sequences in yeast that are unusually productive in creating proteins. The system used simulated annealing to determine the subnets of the protein-expression network that were most active in producing proteins, so that the entire process, not just the DNA-coding aspect, could be identified.
Once these networks are identified, the researchers can then go back to the bench to devise experiments that target those pathways, generating data more relevant to the most active aspects of yeast growth. That data will then be modeled again using Cytoscape. |
