Biological circuitry at Oak Ridge National Laboratory
from the computing-with-microbes dept.
Patrick Underwood writes "A short New Scientist article describes the work of Michael Simpson, Gary Sayler and James Fleming of Oak Ridge, who have modified _Pseudomonas putida_ bacteria cells to produce AND and OR gates. Online at http://www.newscientist.com/dailynews/news.jsp?id=ns9999778"
May 29th, 2001 at 9:34 AM
Relevant links
Here is the website for the group doing this work:
http://www.ornl.gov/ment/main_h.html
Here is a slide show describing this work in more detail:
http://www.ornl.gov/ment/isbc_talk.pdf
Here is other nanotech-relevant stuff happening at Oak Ridge:
http://www.ornl.gov/ORNLReview/rev32_3/brave.htm
This sounds very similar to the work I heard Tom Knight describe in 1997 on what he called cellular logic gates, and later microbial engineering.
http://www.ai.mit.edu/people/tk/ce/microbial-engineering.html
http://world.std.com/~wware/knight.html
Of related interest, a DOE report on nanoscale science and engineering:
http://www.sc.doe.gov/production/bes/nanoscale.pdf
May 29th, 2001 at 11:59 PM
This is of limited significance
Due to the fact that bacteria are at the 1 micron scale and current microelectronics is at the 0.18-0.13 micron level and slated to go to the 0.10 micron (100 nm) level by 2007, this development is of questionable significance. DNA computing (10-100 nm scale) is of some significance because of the amount of parallelism inherent in manipulating large numbers of molecules. Moving up the scale to the "bacterial" computing level seems to be of dubious value.
Fine, its nice that we can manipulate genetic systems so they can "compute". But their speed is limited due to the random diffusion of the molecules (vs. the speed of electrons in microelectronics or the speed of light in optical systems).
It would be better to understand and apply the capabilities where they demonstrate clear advantages. Microorganisms self-replicate, micro-electronics components don't. The advantage is in the self-replication, not in the programmability. Microelectronics and optics have high velocity directed transmission of information carriers (electrons/photons), microorganisms don't. Use the advantages of each system where they provide the greatest benefits.
A hammer is best at pounding nails and a saw is best at cutting wood. Exchanging the tools and attempting to use them for something for which they were not designed does nothing but slow down the construction of the house.
May 30th, 2001 at 4:43 AM
Re:This is of limited significance
Due to the fact that bacteria are [larger and much slower than transistors], this development is of questionable significance.
You're right, typical signal switching times in Knight's work are in the tens of minutes. The real win here is self-replication, so that a large bacterial computer can be extremely cheap. Given that conceptual simplicity was never an important selection pressure, programming these things is likely to be quite a challenge. Eventually most of that complexity will be buried in the depths of a compiler, but it will take a while to reach that opint.
In situations where bacterial computers are applicable, they will be really cheap! Particularly in applications where you don't need an answer immediately, but where you'd like a lot of parallelism, these can make sense. There must be plenty of useful problems of this sort.
John Koza, the genetic algorithms guy, built a big Linux cluster to solve various engineering problems. Most of these problems need not be solved in one day or even one week to be useful. Where he spent millions of dollars on his cluster, a developing nation or an individual could easily afford to spend tens or hundreds of dollars on a large pool of sugary water and a small packet of seed bacteria. I'd put one in my back yard, if it were available today.
May 30th, 2001 at 4:59 AM
Oops, I forgot something
Another thing bacteria have that transistors don't have is direct interfaces to all kinds of interesting environmental considerations. A bacterium may have sensors for light levels, temperature, pH, salinity, electric and magnetic fields, and the concentration of any of dozens or hundreds of different small molecules in its immediate environment. Such things are doable with a normal computer, but generally involve ordering the sensor as a separate piece, providing an analog-to-digital converter, some glue logic, additional connectors and wires, and so forth. The bacterium gives you sensors for free. In some cases the bacterium may also provide free actuators: flagella and motors, or the emission of other small molecules.
April 15th, 2006 at 2:42 PM
i couldn’t understand what a biological cirit is all about. i will like to get an artcle about this.