BIOINFORMATICS GROWS TO KEEP PACE WITH PETABYTES OF SMALL TECH DATA

[Murray Charters <mcharters@xxxxxxxxxxxxxx>]

BIOINFORMATICS GROWS TO KEEP PACE WITH PETABYTES OF SMALL TECH DATA
By Jack Mason
Small Times Correspondent

June 25, 2003 - A microfluidic device can function as a "liver-on-a-chip." 
Forests of nano-pillars can unravel strands
of DNA. These small tech-enabled leaps in life science involve increasingly 
complex interactions, and the result is new
data for scientists working on curing diseases.

Too much data.

In fact, so much data that humans alone cannot possibly make sense of it all.

"It's like a 3-D puzzle in which a piece may fit not with just one, but 
hundreds of others," said Laura Mazzola,
founder of the Nanobio Forum in Northern California. The ability to recognize 
patterns in huge data-bases of genetic
information, said Mazzola, has rapidly outpaced what even the best human minds 
can comprehend.

Today, analyzing how multiple genes function together can produce terabytes of 
data. But as nanotech enables greater
sensing and collecting of data, the info flow could become measured in 
petabytes, or a quadrillion bytes of
information. Muscling such large and complex raw results into useful knowledge 
is the goal of bioinformatics.

Front Line Strategic Consulting predicted last year that the bioinformatics 
business will reach $1.7 billion by 2006.
The market research firm said bioinformatics would grow at a 20 percent annual 
rate while helping shave 33 percent of
the cost, and two years of time, off the drug discovery process.

Stanford bioinformatics professor Russ Altman envisions the day when 
microfluidic systems connected to computers will
fully model an individual cell?s structure and function. Armed with such a 
detailed virtual representation of how a
cell works, scientists will be able to develop novel drugs with unprecedented 
speed and precision "without ever
touching a mouse." The biological kind, he means.

The National Science Foundation (NSF) recently funded a Network for 
Computational Nanotechnology based at Purdue
University. Formed by seven universities, the network aims to give academia and 
industry access to advanced simulation
tools for disciplines including biotech.

Startups like BioForce Nanosciences Inc. in Ames, Iowa, are working on 
nanobiotechnology?s commercial frontier. The
company said its NanoPro system can deposit as few as a thousand biomolecules 
in an array of droplets spaced a few
nanometers apart. Chief Science Officer Eric Henderson said that 1,000 droplets 
can fit in the area of one microarray
well. An atomic force microscope can then "feel" topographical features in the 
samples that signal cancer or HIV.

Public bioinformatics resources include online databases such as the Protein 
Data Bank, a collaboration among Rutgers
University, the San Diego Supercomputer Center and the National Institute of 
Standards and Technology. Web-based tools
such as the National Center for Biotechnology Information's BLAST (Basic Local 
Alignment Search Tool) gene search
engine enable researchers to hunt for patterns across the entire human genome, 
as well as all other known genomes.

Of course, the convergence of bio, nano and computing also encompasses a slew 
of software, open source and commercial.
Stanford's Altman said that all commercial bioinformatics software is only as 
good as the proprietary data and
expertise that may come with it.

Cengent Therapeutics Inc. -- the product of a merger between Structural 
Bioinformatics and GeneFormatics -- is a
privately held company in San Diego focused on proteomic drug discovery. The 
company's relational database, called
StructureBank, can store and compare different 3-D protein structures.

Rosetta Biosoftware Inc. in Kirkland, Wash., produces bioinformatics packages 
for drug discovery research. The
company's Resolver system helps scientists see gene pathways, make predictions 
about how genes interact and identify
patterns. Doug Bassett, Rosetta's general manager, said the new 4.0 version of 
Resolver has a feature that can filter
out the most biologically significant data in a specific experiment to speed 
and focus results.

Accelrys Inc., also in San Diego, offers programs for modeling, simulation and 
analysis of biomolecules as well as a
range of bioinformatics tools. And GeneSpring 5.1 from Silicon Genetics of 
Redwood City, Calif. is one of many software
applications for gene expression analysis. LION Bioscience AG in Heidelberg, 
Germany, markets its DiscoveryCenter as a
total platform integrating drug discovery data, applications and documents into 
a single desktop interface.

Not surprisingly, all this intense computing requires significant hardware 
horsepower. IBM launched its $100 million
Blue Gene supercomputing project in 1999 to help unravel how protein molecules 
are constructed and to advance the field
of biomolecular simulation.

Diseases such as Alzheimer's and Parkinson's are thought to be caused when 
genes build proteins with small errors in
them. But measuring the smallest proteins folding themselves into intricate 
twists could take decades for a current
scientific work-station to process.

IBM's first machine, the Linux-powered Blue G/L, is slated to debut in 2005. 
With 65,000 parallel processors and 16
trillion bytes of memory, the bioinformatics supercomputer is expected to 
process 200 trillion calculations per second.
And with IBM's e-business-on-demand initiative, the company may handle intense 
computational projects such as
bioinformatics as a service or utility, paid for like electricity or water .

SOURCE: Small Times


* * *

----------------------------------------------------------------------
To sign-off Parkinsn send a message to: mailto:listserv@xxxxxxxxxxxxxxxxxxxx
In the body of the message put: signoff parkinsn