AI
Sunday, November 30, 2003 2:27:23 AM
Very good point, minwan0! I'll freely admit that noncoding DNA could be controlling the behavior of coding DNA, but whether the mechanism is passive or active remains to be determined. Although the analogy between DNA and computers has to break down somewhere, if I may, though, I'll try to stretch it just a little bit farther. 
Dr. Rubin said, "That means, in short, that we've found a genetic switch outside of the [protein-coding] genes." (I'm trying to be specific about the functions here.)
Now, computer programs don't operate by generating proteins--they just alter electical pathways, which is much more direct than cellular processes. Still, there are some rough analogies that can be made. A common thing to do in computer programming is to test a value against a limit before performing an action:
number howMuchInterleukin; // this is data
// and this stuff below is code (pseudo-code),
// except for the '4' which is data
//
if howMuchInterleukin is less than or equal to 4, then
make more interleukin;
add 1 to howMuchInterleukin;
done
Here we have a case of data controlling the activity of code, but still in a completely passive way; the code consults the data and bases its activity on it. Noncoding DNA could be operating as simply as this, with the coding DNA "somehow" inspecting it for limits on what it should do. (And nobody seems to know what that "somehow" is, yet.)
Another possibility is that noncoding genes operate through some active mechanism, generating other substances in the cell that inhibit the production of proteins by coding DNA. You might regard this as the way your computer's (more modern) operating systems control how much of the CPU a program is allowed to use. Older Windows and Macintosh systems, for example, had no way of preventing any program from taking over the whole computer. Windows XP and MacOS X, however, run all programs in a time-sharing fashion, in some senses using a different mechanism (noncoding DNA?) than the programs (coding DNA) that are actually getting things done for you. The analogy really starts to break down here because in computers the same exact mechanisms are in fact at work, it's just that the operating system has special access to parts of the CPU that regular programs can't get at. However, the abstract principle is illustrated, which is all I'm after.
So noncoding DNA could be passive, or it could in fact be active (though probably not in the same way that coding is). Or it could be both! I wouldn't be surprised to hear discoveries that some noncoding DNA represents data, while other noncoding DNA operates actively through some mechanism other than generating proteins. The fact that scientists are considering these possibilities and actively investigating them is all that really matters to me, though.
--Nik
Dr. Rubin said, "That means, in short, that we've found a genetic switch outside of the [protein-coding] genes." (I'm trying to be specific about the functions here.)
Now, computer programs don't operate by generating proteins--they just alter electical pathways, which is much more direct than cellular processes. Still, there are some rough analogies that can be made. A common thing to do in computer programming is to test a value against a limit before performing an action:
number howMuchInterleukin; // this is data
// and this stuff below is code (pseudo-code),
// except for the '4' which is data
//
if howMuchInterleukin is less than or equal to 4, then
make more interleukin;
add 1 to howMuchInterleukin;
done
Here we have a case of data controlling the activity of code, but still in a completely passive way; the code consults the data and bases its activity on it. Noncoding DNA could be operating as simply as this, with the coding DNA "somehow" inspecting it for limits on what it should do. (And nobody seems to know what that "somehow" is, yet.)
Another possibility is that noncoding genes operate through some active mechanism, generating other substances in the cell that inhibit the production of proteins by coding DNA. You might regard this as the way your computer's (more modern) operating systems control how much of the CPU a program is allowed to use. Older Windows and Macintosh systems, for example, had no way of preventing any program from taking over the whole computer. Windows XP and MacOS X, however, run all programs in a time-sharing fashion, in some senses using a different mechanism (noncoding DNA?) than the programs (coding DNA) that are actually getting things done for you. The analogy really starts to break down here because in computers the same exact mechanisms are in fact at work, it's just that the operating system has special access to parts of the CPU that regular programs can't get at. However, the abstract principle is illustrated, which is all I'm after.
So noncoding DNA could be passive, or it could in fact be active (though probably not in the same way that coding is). Or it could be both! I wouldn't be surprised to hear discoveries that some noncoding DNA represents data, while other noncoding DNA operates actively through some mechanism other than generating proteins. The fact that scientists are considering these possibilities and actively investigating them is all that really matters to me, though.
--Nik
