Now on the ‘Net: Life’s source code

Bioinformatics, the management and analysis of biological information, is today roughly where the Internet was 10 years ago. With the nearly finished Human Genome Project, miraculous transformation of HIV from death sentence to manageable disease and development of “biochips,” this nascent IT sector has only just reached the starting line. (An overview of the HGP is available here.)

Storing, searching, analyzing and sharing terabytes of data is what bioinformatics is all about. That’s good news for the computer and networking industries, though so far most of the action has been confined to a relatively small number of government, academic and commercial research centers.

That situation might change soon. Although the biotech industry has not come close to delivering on its promise, Moore’s Law is finally coming into play, and solutions soon will begin trickling down to physicians’ offices.

Personalized medicine is bioinformatics’ Holy Grail. We’ve known for some time that a specific drug might work well for some people while causing harmful side effects for others. Instead of the Federal Drug Administration rejecting a drug because 1% of patients suffer serious side effects, we will be able to produce custom-made drugs for different groups and, ultimately, individuals.

This is where biochips come in. Biochips are small devices that can be used to perform thousands of tests simultaneously on a single sample. Applications range from testing for specific genes to pinpointing disease types to monitoring treatments.

That’s not to suggest biochips ever will replace research centers. On the contrary, biochips and research centers complement each other: Small, affordable field devices would collect data on entire populations from which bioinformatics centers would design more useful biochips.

The challenge for bioinformatics is making sense out of genomic and proteomic data that is growing by leaps and bounds. Living systems contain thousands of genes and proteins, many of which interact in subtle ways. Finding order in seemingly chaotic data calls for more storage, processing and transmission capacity.

The first response from major research centers and pharmaceutical companies was to purchase expensive supercomputers. But often what’s needed is the ability to perform many small processing tasks simultaneously. Bioinformatics centers increasingly are turning to storage-area networks, server clusters from companies such as Hewlett-Packard and Linux Networx, and even grid computing from companies such as Platform Computing and United Devices.

Understanding underlying disease processes is key to finding effective treatments and eventually cures. Helping to assemble all pieces of the puzzle will be one of the Internet’s finest accomplishments.