Install

Get the latest updates as we post them — right on your browser

. Last Updated: 07/27/2016

DNA Could Hold Key To Faster Calculations

LOS ANGELES -- Computer scientist Leonard Adleman watched as associates in a research lab toiled with DNA, the chain of molecules that make up the genetic code of living things.


Adleman was amazed at the intricacy of the DNA strands, which contain the blueprints for the assembly of everything from tiny viruses to the biggest mammals. And he was struck by how similar the cutting, splicing and copying of these strands by the laboratory were to the manipulations of numbers he performed with computers.


Then came the light bulb. Could DNA somehow be harnessed to tackle calculations, perhaps more efficiently than man-made machines?


The University of Southern California computer-science professor devised a test using a batch of synthetic DNA molecules to compute a relatively simple and common mathematical problem.


That problem, involving a theoretical travel itinerary, is still the only one that has been solved using his DNA technique. But as soon as he published his results in November in the journal Science, some of his colleagues took notice and began trying to build on his work.


Instead of the chains of simple numbers used by silicon-chip-based computers, or letters used to make words in human language, the basic units encoding DNA's information are four tiny molecules called adenine, cytosine, guanine and thymine. They are abbreviated A, C, G and T. These four building blocks can fit together to form more complex numbers or words, which then combine to make larger words.


The sequences of A, C, T, and G molecules are "kind of like LEGO pieces that stick together,'' said Don Beaver, a cryptographer and assistant professor of computer science and engineering at Penn State.


To test his theory, Adleman used a form of the Traveling Salesman problem, a simple puzzle which requires picking the shortest flight route to connect seven cities with no more than one stop at each. He gave each trip segment a flight number represented by a piece of DNA made of a particular sequence of those A, C, T and G molecules. Mixing those pieces together in a solution then allowed them to stick together and form various new combinations, like stringing words together to make a new sentence.


To read the answer, a separate chemical reaction has to be used to determine the identity of each little group of As, Cs, Ts and Gs along a strand of DNA.


For problems with one unique solution, a series of these tests would pick out the strands that hold the right series.


For more complicated problems, Beaver explained Tuesday, a sort of chemical Velcro would have to be used that would stick to one particular combination or word and pull out the DNA strands that start with that word.


Because it relies on simple chemical reactions rather than the movement of electrons on silicon chips, the process is more energy-efficient.


DNA computing should yield some solutions a million times faster than electronic computers, because billions of DNA molecules could carry out billions of operations simultaneously, even though each individual step might be fairly slow.