DNA On A Chip and The Birth of Computational Biology

DNA On A Chip and The Birth of Computational Biology

by : Aria Ratmandanu 

          Computer scientists engaged in DNA sequencing will not simply pack up their bags and go back home in 2005 when the project is done. This is because the Human Genome Project is just the beginning of an entirely new science.

        It’s turning biology into an information science. Many biologists consider the acquisition of sequencing to be boring. But from a computer science point of view, these are first-rate and challenging algorithmic questions,” says Richard Karp of the University of Washington, one of the leading computer scientists in the country.

       Computer science first invaded the world of biology in 1983 when Russell Doolittle and his colleagues rocked the closed world of molecular biology by making a major biological discovery by simply reading computer printouts. Without performing a single experiment, Doolittle was able to find a similarity between two dissimilar proteins involving different areas of biology: the sis cancer gene and a cellular growth factor. He and his colleagues noticed that the DNA sequence found in this particular type of cancer was also the same DNA sequence involved in cellular growth, thus showing that cancer genes created abnormal growth in cells. Biology was not supposed to be done this way.

        Robert Cook-Deegan of the National Academy of Sciences asked the rhetorical question: “Why should he be able to publish a major discovery that came from just sitting at a computer terminal? That wasn’t biology, was it ?

       This dramatic discovery heralded the beginning of using computers to spot patterns in DNA sequences rather than getting one’s hands dirty with test tubes of proteins. “The intrusion of computers into molecular biology shifted power into the hands of those with mathematical aptitudes and computer savvy,” Cook-Deegan notes. “A new breed of scientist began to rise through the ranks, with expertise in molecular biology, computers, and mathematical analysis.

         In the past biologists learned about life by analyzing the interior of living specimens (i.e., in vivo). In the last century, they learned to study life in glass (i.e., in vitro). In the future, they will study life via computers (i.e., in silico).

        What will the sequencing process look like in 2020 ? Will we have thousands of acres devoted to housing monstrous computers and robot factories that sequence people’s DNA ?

        Probably not. Just as the future of computer technology lies in miniaturization via the microchip, many scientists feel the future of DNA sequencing will be the “bio chip” and the “DNA chip,” in a grand merger of the computer and biomolecular revolutions.

      The bio chip is a microchip designed specifically to perform “homology” searches between similar human and animal genes. This bio chip is enormously useful for biologists, because if a certain genetic sequence in an animal is already known to control a certain protein, searching for its counterpart, or homologue, in humans reduces the guesswork involved in identifying unknown human genes. In the future, in lie with Moore’s law, the bio chip will eventually take over the business of DNA analysis.

        A primitive bio chip already exists—it is a quarter of an inch across, contains 400,000 transistors, and “is the most complex chip that the Jet Propulsion Laboratory at Caltech has ever designed,” according to Leroy Hood. It is about 5,000 times faster than a Sun Sparestation 1. When instructed to identify a 500-base sequence among 40 million bases, the Sun Sparestation computer took five hours, while the chip took only 3.5 seconds.

           Scientists are now perfecting a DNA chip as well, a microchip that can almost instantly screen a person’s DNA for selected genes. DNA chips, which will soon begin to enter the marketplace, can test for HIV, cancer, and thousands of genetic diseases within a matter of hours. This new diagnostic tool may revolutionize the $17.5 billion diagnostics industry.

           With the advent of the DNA chip, Gilbert’s dream of “personalized” DNA records that contain all our genes is no longer pie in the sky. Already, several start-up biotech companies are racing to read our DNA by scanning them onto microchips. The merger of computers and molecular biology on a DNA chip may signal a new era in cheap, rapid genetic screening.

          To the naked eye, the DNA chip seems to be rather unremarkable. Only the size of a fingernail, it looks very much like the microchip that is used in most PCs. But under a microscope, you would see a most unusual pattern. With the same photolithography techniques used to etch microscopic grooves in tiny transistors, scientists use a template to etch the outline of DNA strands corresponding to a particular set of base pairs. By washing a solution of DNA over these templates, those sequences which remain are those which fit precisely into each probe.

         The trick is that the only DNA strands which stick to the microchip are those which precisely fit the template pattern etched onto the chip. All other strands are washed away. A laser then makes some of these sequences fluorescent and a computer makes the final identification.

         Already, the Affeymetrix corporation is marketing a chip with 65,536 probes etched onto it,  each probe being the template for eight base pairs. “We have actually produced a prototype chip containing a million probes,” claims Robert J. Lipshutz, the company’s director of advanced technology. Affeymetrix has already succeeded in placing all the genes for HIV on a DNA chip, which can accelerate AIDS screening dramatically.

         The potential for the DNA chip, which squeezes an entire DNA laboratory onto a single chip, is enormous. Already, one can use it to screen for the notorious p-53 mutation, which is implicated in over half of all cancers. Cystic fibrosis, which comes in any of 450 different mutations, can be screened by the DNA chip in a few hours at a cost of only a few dollars. (The traditional process of identifying these cystic fibrosis genes is quite expensive and takes at least a week.)