roy@alanine.phri.nyu.edu (Roy Smith) (10/30/90)
A high school junior just introduced himself to me, looking for a
mentor for a Westinghouse project. It sounds like what he wants to do is
something in computational biology. He mentioned things like RNA folding
and molecular modelling. He's just learning C and is taking a course which
I think he said is called Biological Research but sounds like it's really
"Let's Do a Westinghouse Project, 3 credits per semester, by permission of
the instructor"
I'm not sure how to approach this. From the 3 minutes I got to
talk to him, he seems pretty bright, but I'm not sure how serious a project
even a really smart high school junior can be expected to tackle in a
year's worth of effort. When I was a junior or senior in high school, a
really cool computer project was writing Hunt The Wumpus in BASIC; clearly
that's not going to win any Westinghouse awards. Didn't I hear somewhere
that Ray Lau wrote StuffIt for the Mac while he was in high school? Have
things changed that much in 15 years, or is Ray just a lot smarter than me?
I certainly don't want to just think up a year's worth of busywork
for him to do, but I also don't want to end up with a project that's too
big or too hard to get a handle on. I'm not even sure how much of the
planning I'm supposed to be doing anyway; I always look at the projects
Westinghouse winners do and think, "no way did a high school kid conceive,
plan and excecute that on his own" and don't want that to happen here. Any
ideas on how I should deal with this?
I guess the big question is, it is reasonable to expect that a kid
just learning C now could possibly, in a year from now, produce some
useful, impressive, or just plain interesting body of work in computational
biology?
--
Roy Smith, Public Health Research Institute
455 First Avenue, New York, NY 10016
roy@alanine.phri.nyu.edu -OR- {att,cmcl2,rutgers,hombre}!phri!roy
"Arcane? Did you say arcane? It wouldn't be Unix if it wasn't arcane!"mmm@cup.portal.com (Mark Robert Thorson) (11/04/90)
A reasonable application of computers for an introductory C programmer would be the collection of statistics. I remember how impressed I was to learn about the technique known as the evoked potential. Under computer control, you flash an input stimulus, then begin sampling from your electrodes. On a single trial run, the data looks like random noise, but collect 100 trials and average them out, and you see a distinct characteristic wave in response to the stimulus. The stimulus can be almost anything. The experiment I saw used bit-patterns flashed onto the frame buffer of a computer-graphics display. The response is a bit sticky. If I'm correct, the GE show forbids venipuncture of warm-blooded animals. I don't know what the regulations are on surface electrodes, but I'd guess that recording from humans and mammals is verboten. You can do anything you like with invertebrates. I suggest some sort of measurement of the visual processing of a snail or cockroach. Develop equipment to take brainwaves, then do evoked-potential work. For example, you clamp down the animal by gluing it to a styrofoam block. Your stimulus is a dark/light edge moving from left to right. This is done by having a black card under the tension of some rubber bands. The computer flips a bit, and a coil releases a catch so the card flies past the animal's visual field. Immediately when the command to release the catch is given, the computer begins recording brainwaves at 1/100 second intervals. Average 100 trials, and you should have some sort of interesting result. Sound is another easy stimulus variable. The computer emits a sound from its loudspeaker, and you capture a block of data from an A/D converter. Average 100 trials, and you should get some sort of characteristic wave. Some prominent scientists believe you can measure IQ by recording the evoked potentials observed on scalp electrodes applied to human subjects upon whom certain test images are flashed. I can provide references if needed.