The next generation of computers may not be housed in plastic and metal, but within test tubes and perhaps even in the human body itself. A DNA computer could be thousands of times smaller than any silicon chip currently being made. And while DNA computers may never be at the heart of the next must-have gaming console, they might be in the gamers of the future, regulating body chemistry and treating illnesses like diabetes.
A research team at SUNY Geneseo is working to help compile a library of DNA strands suitable for such technologies.
The iconic image of DNA is a double helix; a twisted ladder with nucleotide base rungs. These bases are Adenine(A), Thymine(T), Guanine(G) and Cytosine(C), and they fit together in specific pairs. Base A should only bind with base T and base G should only bind with base C. This is known as Watson-Crick base pairing, and it's exactly these rules that make DNA computing possible.
Regular computing works in binary. The computer reads electronic pulses that represent the digits 1 and 0. A DNA computer would read information in A, T, G and C. DNA computing would be similar in practice to a type of computing called parallel computing, where computers are linked to one another and all work on different parts of one problem at the same time. It's not that DNA can solve a problem that much more quickly than silicon; it's that DNA is so small that millions of possible solutions can be introduced, and because of base pairing, only the right answers stick.
But Senior Lauren Wood and Sophomore Arunima Ray, both students of biochemistry and mathematics, aren't interested in strands that pair up right. They're interested in mistakes. If something called cross-hybridization of DNA strands occurs, the bases won't match up correctly. Cross-hybridized strands don't show Watson-Crick pairing. Sometimes A binds to C, or bulges form in the strand, disrupting the helix structure. If such strands are used for processes like DNA computing, the results are unreliable.
For DNA computers and technology that uses DNA in similar ways, cross-hybridized strands need to be recognized and eliminated. Wood and Ray are working with Wendy Knapp Pogozelski, Professor of Chemistry, and Anthony Macula, Professor of Mathematics, to test a system that can detect cross-hybridized DNA strands. Their technique uses the fluorescent dye SYBR Green I, which causes cross-hybridized strands to fluoresce, or light up. Finding out which strands are cross-hybridizing allows the researchers to create a DNA library, a database of strands that don't cross-hybridize, that would produce reliable results if used. One of the major benefits of this new technique is that it allows scientists to test many different strands at once, accelerating the construction of a reliable DNA library. Previously, scientists could only compare two strands at a time.
What Wood and Ray are doing is called a proof of principle. The technique has successfully identified known strand cross-hybridizations, and now it needs to be proven again by testing the technique in larger, randomly assembled pools of DNA. If this technique is further validated, it may bring us one step closer to a new world of computing. Wood and Ray recently presented their research on "Detection of Cross Hybridized Strands Within Pools using SYBR Green I Fluorescence" at the Joint Meeting of MAA and AMA in January.