Friday, October 01, 2004

When experiments go wrong ...

... they can have unfortunate consequences, like the one above, after I tried to get some DNA to do some experiments. For some reason, the fellow in the white shirt really didn't want to have a big needle stuck into him ...

I've actually started doing some experimental work, which is basically going to consist of trying to glue some pieces of DNA together. My impressions so far:

- you work with really, really, really small amounts of material. For one of the mixtures I had to prepare, I had to prepare 0.5 microliters of solution. To give you an idea of how little that is, take the contents of a soda can and divide that into about 300 equal parts. Take one of those parts and divide it into 1000 parts. Now take one of those parts and split it into two. That's about 0.5 microliters. The amazing thing was that I didn't actually need any fancy machinery to do this -- there are entirely mechanical, manual pipettes that you can use to measure out something that small.
- A lot of the experimental procedures work simply because of the sheer immense numbers of molecules involved ie even though the chances of two molecules reacting the right way are pretty small, if you have enough molecules, the right reaction will happen often enough that you can actually get something useful done. To give a concrete example: the work I'm doing basically involves taking 2 pieces of DNA, mixing them together with molecules that cut the DNA strands at very specific spots and then trying to glue matching pieces back together simply by mixing them together in a test tube with other molecules that act like glue. There are lots of ways this can go wrong: the DNA has to get cut in the right places, the glue molecule and the two bits of DNA to glue together have to be in the same place at the same time for long enough that the reaction can happen, the right bits of DNA have to be next to each other to get glued together etc but somehow it works often enough to be useful.
- There's lots of room for automation. As cool as pipetting is the first few times, by the time you're halfway through the 10-step procedure for the first stage in your 10+ step process, for the first solution of the 10 solutions you're making, you're pretty tired of it.
- Temperature and timing are really important. All the incubation rooms and fridges are set at very precise temperatures - 37 degrees celsius, 4 degrees celsius etc, and kept at that temperature by lots of expensive machinery. Letting a reaction go on for a bit too long or too short can totally screw up your experiment and then you face a bunch of trial and error trying to figure out what went wrong.

All in all, though, I'm enjoying it so far. Let's hope it stays that way.


Blogger Corey said...

Dude! You're like Greg, from CSI. That is, before he went to the field this season...

4:23 PM  

Post a Comment

<< Home