Sorting through DNA

An Anonymous Coward sent in this note: "There's an interesting article at Forbes.com discussing the issues scientists are finding in analyzing and sorting the DNA info now that it's been sequenced. It seems that the functional and useless DNA is all mixed up together. I like this line: "If you want an analogy, think Microsoft Windows: a big kluge of code containing long stretches of old subroutines and dormant bugs and all sorts of other stuff left over from past generations. It shouldn't work, but usually it does.""

Comments

What Arrogance

[destine]

I find it interesting how these scientists have "cracked the code" of the human genome and yet none of them can produce anything truly useful from this knowledge. This is just the first step in a very long process. We might have figured out which of our genes are active and which aren't, but we have almost no idea what activates them. We're barely able to produce differentiated cells from stem cells, and as far as I know, we've only got one company thats figured that out, and they aren't sharing. We seem to be pretty arrogant to say that this is buggy code when we don't even know what we're looking at. We have gene sequences that produce proteins. This is a like a data segment in a program. What are the triggers? Where is the processing? How does it all work? No answers? Thats because we don't undrestand it, and until some guy can go through and start changing fundamental bodily processes, and I don't mean making a third arm or giving someone, I'm talking about creating entirely new transhuman processes into the body, then none of these articles will impress me. Its just plain arrogant.

Re:What Arrogance

[sphyg]

Hey, we're working on it. Okay?

Re:What Arrogance

[BobGregg]

> I find it interesting how these scientists have "cracked the code"
> of the human genome and yet none of them can produce anything truly
> useful from this knowledge.

I hope you're just trolling, because it's difficult to imagine how someone could refer to the foundation of all future genetic work as "[nothing] truly useful". Why not just say Darwin's theory wasn't "truly useful" either? After all, it didn't solve any human problems either, at least not directly. If the scientific revolution of the past century has taught us anything, it should have taught us that understanding and knowledge of reality is *always* useful, even if there isn't an immediate application.

And besides that, useful (functional) applications of the genome knowledge *have* already appeared. Aside from the obvious (genetic therapies, recents insights into Alzheimer's, and on and on), the anthropological knowledge revealed by what's in the genome is giving a fascinating window into the history of human evolution. Consider: we've found that two people of different "races" commonly have less genetic differentiation than two people of the same "race". Given the history of prejudice, racism, genocide, and other horrors that one population of men has inflicted on another in our history for no better reason than the color of their hair or the shape of their nose, the genome is giving us an education in ourselves that we've desperately needed for centuries. If that isn't "useful" knowledge, then I don't know what is.

Re:What Arrogance

[destine]

That was bad wording on my part. Yes this is fundamental and much much needed information. My main point is the misinformation that seems to be plaguing the mass media seems to say things that are misleading or not even true. With headlines like "DNA As Buggy Code" where they equate a working delicate and so far not well understood strand of DNA to a bloated buggy untested piece of windows code. And you might disagree, but this is just a really small step in the great scheme of things to come. I don't think it should be blown out of proportion. People are going to start thinking mad scientists are out there changing around genetic code willy nilly without any forethought and thats not whats happening.

We don't have a buggy piece of windows code here. We have an elegant masterpiece of genetic material that our only real knowledge of so far consists of an overhead map of which we only the rudiments of knowledge. We've still got a long way to go. And while I agree completely that "understanding and knowledge of reality is *always* useful, even if there isn't an immediate application," I also think that we as scientists and people need to step back a moment and see that our crowning achievement with DNA will only come once we've had sufficient time to study and understand it more. And this is what the human genome project is going to allow.

These "functional applications" from the genome have been, at most, informed but lucky stabs in the dark. And generally, cures for genetic diseases have come more out of the study of cells and living processes than anything. We understand that familial Alzheimer's *may* have something to do with chromosome 19. But biologically we *know* that we can reduce the likely hood or increase the longevity of brain function by reducing the breakdown of acetylcholine in the brain. The current genetic therapies are either extremely slow tedious and nonpermanent or are extremely dangerous as producing a proper vector for gene replacement in living cells is still much in its infancy and we still don't know enough about how it works to really do this well.

And besides this specific article having the few bad analogies, its no different then any of the other news pieces on this same thing. These "anachronistic sequences" that it speaks of are actually our past and there is a lot of good that can come from that information. "It shouldn't work, but usually it does," is one of the most arrogant statements I've heard on the subject. Having so little knowledge of how it works how can we say that it shouldn't work.

This article just left a bad taste in my mouth and wasn't nearly as useful as some of the others that I've read recently. I'd give it a C- if I had read marking pin available.

Re:What Arrogance

[joto]

If something is a mess (buggy code), then there has obviously been random processes at work (i.e. random features, random bugfixes, etc...). Evolution is an example of a random process (it is not entirely random, but there certainly are random elements within it). On the other hand, if something is elegantly carefully crafted, that would mean someone intelligent (i.e. God) would have designed it.

So, if you believe that humans are a product of genetic evolution (which is more or less the only scientific theory at the moment), and you believe that evolution is somewhat random, it would be very hard to explain why the genetic code should not be a mess.

There are examples within programming, that potentially could teach us a lot about how the genetic code works. First let's look at the experiments of genetic programming. Genetic programming is a method of evolving programs by letting programs combine and mutate, where combinations of programs are allowed to "mate" depending on their "fitness" and some randomness. The resulting programs, as you can imagine are not exactly elegant masterpieces, but usually horrendous stupidities, that does things in an awkward roundabout way that just happens to produce the right result (or something that is right most of the time). But sometimes it works, although I haven't heard of many practical uses for the technique.

Genetic algorithms is a similar idea, but is usually applied to optimization problems. As opposed to genetic programming, genetic algorithms actually have been used succesfully for several real-world applications. The idea is the same as that behind genetic programming, but here we use evolution to tune some parameters to find the "best" function. There is no guarantee that the resulting functions after say 10 000 generations is especially good, but with good tuning of the genetic parameters the technique has been found to be useful for some specific problems.

The human immune system does from what I understand look very much like the result one would expect from a genetic algorithm. There are lots of individual parameters that must be fine-tuned, and they interact in non-trivial ways. The resulting mess works, but one could hardly call it elegantly designed, as it is a result of evolutionary processes, not divine design. This very much speaks in favour of genetic algorithms as a model of evolution as it occurs in nature.

When we look at the genetic code, we must (if we believe in the theory of evolution) believe that what we are looking at is the result of billions of years of evolution. While we may not totally understand the process of evolution, we believe that it's results would be mostly like the results of genetic programming or genetic algorithms. Of course, nature might be better at it than we are, or perhaps nature is flawed. Only research can answer basic questions like these.

Conclusion: It is not arrogant to say that the genetic code is a mess. We don't even have to look at it, it is after all what we expect. It is possible that further examination might find that, in fact, the genetic code is well-designed and understandable. I highly doubt that, but we can never rule that possibility out (not even after 1000's of years study, as it can be hard to understand someone elses design, in a very alien programming language).

Hold on a minute...

[meridoc]

Think of "cracking the genome code" not as a claim of understanding something, but as meeting a new species.

Here's an analogy (just play along).

We've met this new alien species (let's call it "George the genome"), which has been around for a while. We're finally able to bring him home and study him. So we examine him and we've finally described how he looks and what he's made of. But we don't know how George works yet. Somehow, he produces proteins and takes in information and can synthesize stuff. We don't yet know how to talk to George. We know that if George is exposed to particular chemicals, he'll react in particular ways. If we try to mimic things that we know George will react to, will he react in the same way or will he ignore us or go nuts? Are there other things that will "trigger" George to react?

We have previously discovered (often by accident) that we can trigger some reactions when playing with the genome. By describing the human genome, we can now start to work on figuring out how it works. It's just the beginning.

Re:Hold on a minute...

[Cheese_isgood]

Using your analogy: Scientists have found that George the genome should arrive at Earth in the year 2003. True by that time we might only be able to communicate using pig latin or sign language to George, but we're still closer to understanding George than in 10,000 BC when we could barely communicate to ourselves. The human genome is going to be understood sooner than you think. Before you know it, eating green beans the size of a chopstick will be common and eating peas the size of golf balls would not seem unusual either. Prepare "meridoc" the human genome (George) is coming to a laboratory near you!

God was a hacker...

[tomzyk]

So God hacked a little while coding us. Big deal. Are YOU gonna tell him he's got some buggy and inefficient code in there?

Wolfram and Complexity

[Elbelow]

It seems that for the last eight years Wolfram has essentially been running his company by remote control, working all night, every night, on a new kind of science, which is the name of the book he will soon publish to describe it.
Wolfram's new science propounds an extraordinary idea: With a few basic objects and a few rules of behavior--run a few hundred million times--he believes it is not only possible to create structures of great complexity, but the universe itself, including its vast regions of apparent chaos.

While Stephen Wolfram is undoubtedly a very clever man, I seriously doubt whether he is the first or only person to investigate these kinds of concepts.

Correction to above:

[Mossfoot]

When I say "here's how he described evolution" I meant my teacher, not Darwin :) Bob Vila isn't THAT old...

What ignorance.

[theoriste]

Nobody's claiming that the genome is "done". You haven't looked at a basic molbio textbook recently, have you? Some genetic information we know a lot about. Some we know very little about. You said we have no idea what activates genes. Wrong. What are the triggers to gene activation and expression? There are many. Look up mechanisms for signal cascades due to things called "hormones". Look up "protein kinase C". Etc. Where's the processing? Genes get transcribed in the nucleus. The mRNA (as a for instance) gets processed and then transported to the cytosol, then translated to proteins. How does it all work? We're well on our way to understanding that. The elucidation of the genetic code is just the first step. We have the Rosetta stone elements of the expressed genes on one side, and the genome just published on the other. There'll be plenty of surprises for years to come as we fill in the missing pieces, but we are NOWHERE near as backward as the picture you painted in your posting.

The Red-Green Analogy.

[Mossfoot]

This doesn't really surprise me, and here's why.

When I took a History of Science course in university, we eventually got to Darwin, and here's how he described evolution. But first of all, I hope you guys know what the Red-Green Show is. If you don't, hopefully you'll still get the idea.

People tend to think of the human body to be like something built by Bob Vila. When there is a percieved need (a new table for the house, for example), it is done from scratch, it is something crafted, measured, and precicely built from an expert resevoir of knowledge. This is not the case. It is much more like Red-Green, where the Handyman has a percieved need, and instead he looks around in his room full of junk and figures out what he can duct-tape together to make a passable subsitute.

So it surprises me little that the DNA is such a mess, we're just duct taping together all the useful junk in our geneitc code.