Slashdot Mirror
A Map to Nowhere?
Aruges writes "It seems as if there some serious doubts about the value of the human genome map. The main thrust of the problem is that since there are far fewer genes than once thought, the old idea of "one gene, one protein" has fallen by the wayside. The upshot of this is that it may be several decades before we see any benefits, if we see them at all.
Check this story on Spectator for more information."
I do think I would add a few comments though. To start with, it seems that the artical really didn't credit how much progress has been made, and it is obvious that the artical is not being written in context of the project. Specifically, I would point out two things:
First, not to many years ago, there were numerous articals that discussed how difficult the project would be... and giving credit to how far ahead of the predictions the Human Genome Project is really.
Secondly, IMHO, no one has given enough credit to how much Moore's Law has really played in the project. It was calculated the length of time to decode the human gene, and deemed impossable. It was only in light of Moore's Law that serious work began. Now, it's at a level that outpaced most estimates.
It's only because of how far we have come that people say things like "map to nowhere." Keep in mind, when people tried to find alternate trade routes to India by sea, finding America was probably considered "a map to nowhere."
In the service packs!
---- "If we have to go on with these damned quantum jumps, then I'm sorry that I ever got involved" - Erwin Schrodinger
Well, the entire genome can be fitted on a CDROM. That isn't very much data at all. Are we really saying that the human body is no more complex that a copy of Windows 2000?
Obviously, it is.
Not so "obviously" at all. A complex-looking object can easily arise from simple rules and/or simple building blocks.
When I build a house, I don't have to specify where every brick is laid. I just have to tell you how bricks fit together to make walls, and where I want the walls to be.
The Mandelbrot set is another example, for the math geeks among us. It looks infinitely complex, and it _is_ infinitely detailed, but the algorithm that produces it can be stated in one sentence.
In summary, the amount of data needed to describe the human body could conceivably be quite small.
Overall this article seems to be more a demonstration of the ability of journalists to blow scientist's claims out of proportion than the ability of scientists to make important, but incremental, discoveries. I think the article tries to be provocative, and it's not completely off-base.
However, there are number of factual details that the author leaves out here:
The number of genes predicted by these programs varies wildly (Incyte claims 100k+, Celera says ~30k). This is because these different organizations use different software packages to find the genes. But the reality is that no-one really knows yet. No one has actual studied all these genes, yet. No one has purified the actual protein molecules that they are supposed to encode. Very few (less than 1000) have been extensively studied, and in almost every one we've learned more about what makes a gene.
The sequencing of the genome is a hugely important achievement. But, as other have said, it is just a first step. It gives us the substrate upon which many many years of research will be based. It's clear that there is a lot of work left to be done. Already many research labs are moth-balling their DNA sequencers and replacing them with NMR magnets to examine the 3D structure of these gene-products. The protein folding problems remains one of the most studied, unsolved questions remaining in bioinformatics.
Hope this was somewhat useful.
-c"If you are an idealist it doesn't matter what you do or what goes on around you, because it isn't real anyway."-R.P.W.
A lot of optimism assumed that a single gene had a single function, for all time through the life of the cell. But as one would expect with a biological system, things are far more complex. Geners are "turned on" and "turned off." Multiple genes interact indirectly.
The key is that the cell is an emergent system. It exhibits extremely complex behavior as a result of vast numbers of interactions of simpler parts. Thus we may never find a "gene" that "codes for" the shape of a nose. The fact that a nose arises at all from a bunch of protein specifications is itself a clue that things are extremely complicated.
The decoding of the genome will indeed be extremely valuable. But it won't "solve" biology anymore than the understanding of the laws of gases "solves" the weather forecast problem!
The only good weather is bad weather.
I am writing a paper for Molecular Biology about this right now interestingly enough. For a long time we have had little information about bitter, sweet, and umami (monosodium glutamate... meat! (well mostly)) taste receptors.
But one day someone found that a genetic mutation at a specific allele can cause changes in a mouse that effects perception of a specific bitter taste. When they researched the area around the mutation with the genomic database, they found an entire bitter receptor. When they searched for similar bitter receptors across the entire genomic sequence, they found an entire array of bitter taste receptors! This goes for humans too. With so much less work, we've found a whole class (the T2R) of bitter taste receptors.
Now you all may say "WHO GIVES A FUCK?!?", but this is actually important. Since we never knew how sweet receptors worked, we've always guessed about what compounds will substitute for sugar. For example, all the artificial sweetners that I know of were found accidentally, with a chemist tasting the substance during lunch or dinner after working with it for something unrelated. Now that we're gaining more information about receptor function we could conceivably find the perfect artificial sweetner... BILLIONS OF DOLLARS.
Besides, if you can't see by now that finding new receptors and quick searches for similar (consensus) sequences in the genomic database isn't going to herald all sorts of scientific advances, you're pretty dense... While it's true that there are alot of different ways the genetic code is used (substitution of polyadenylation sites in leukocytes, different intron splicing in cochlea frequency detectors), we know about alot of them, and with more analysis we will continue to learn more about ourselves even faster than we ever have before.
Genomics might be that kind of problem.
I think the article only confirms that the author doesn't know much about genetics. No geneticist (that I've even spoken to, at least,) believes that a list of genes equals a useful API for making/modifying humans.
The clones we made of sheep, mice, and other animals resembled the products of buggy code made by lazy programmers or those forced to write shit by insane business models (remember the sig, "it compiles! Ship it!"). We didn't realize the significance of the slow, steady process of genetic replication within the embryo.
Exactly right. Cloning is a "neat hack" in the Computer Science sense: I don't know what X does, but I know enough to make a copy of it. No one involved in cloning claims they understand all the processes involved in growth... they just have used a bunch of tools to provide a proof of feasibility. Think of them as writers of "bit copiers" (for those that remember the old days of floppy disk duplication.) Ask them if it's an exact copy, and the better of them will say "I'm not sure, but it seems to work so far."
Likewise, we hurled gazillions of dollars at the genome project, in private and public searches. Why were the gazillions hurled? Because of the notion that we could find nice, patentable pieces of genetic code, controlling various physiological processes.
Now that we realize we have got a map to nowhere, lets table the whole deal until we understand more about the operation of genes.
Yep, "map" is a dumb term. I'd prefer to think of it as having the object code to an operating system and its associated applications, running on a processor that we don't have the spec for. Some crashes we can cure (they occur in application code that is clearly fixable,) but most problems are due to interactions amongst parts of the system.
Cancer is similar to the infamous Blue Screen of Death. Yes, it's obvious that something bad has happened, but we aren't going to find a line of code that says BlueScreenOfDeathNow(). It's an emergent property of a complex system.
I am all for scientific research but I worry that further pushes down this line of inquiry will be driven by the profit motive, not any kind of medical or healing motive.
You're more of an optimist that me. I reckon most easily curable deseases with be cured pretty soon. The bulk, however, will fall into the category of: "system crashes after 70 years uptime. no solution other than complete redesign."
One thing that is not at all well conveyed by the article is that people have known that the "one gene/one protein" view is oversimplified for quite some time. The concept of splicing variants, i.e. that a single gene produces a variety of related products, is something that my coworkers take as being so natural that it doesn't even bear mention. Yes, it's true that the textbooks may need to be rewritten, but that's because textbooks are always decades behind the cutting edge research.
In any case, people have already been working out ways to take advantage of the genomic data without needing to figure out in advance exactly how it is processed to produce proteins. That's because figuring out the exact splicing points is very tough, and people wanted to use the data before it was completely annotated. Thus the techniques they've been establishing are already well suited to dealing with multiple proteins coming from a single gene. It's a bump, but nothing like the drastic problem presented in the article.
There's no point in questioning authority if you aren't going to listen to the answers.
I see plenty of people dismissing the parent as a troll and wonder if they are being fair or not. True there has been more than the average amount of trolls written in the past couple days or so. But I think that Christian Soldier has a valid point here. The human genome is complex, yes, but I think that it is correct to say that "who we are" cannot be fully attributed to the genome. There is an awful lot of complexity to our very existences (and I'm getting a bit metaphysical here so I apologize.)
I realize that Slashdot has plenty of atheist/freethinker types. Hell, I myself am an agnostic, a lapsed Catholic actually (which would make me more hated than atheists by some branches of fundydom! But that is a different topic all together, I would say.) But I do think that those who call "Christian Soldier" a "troll" are trying to discredit his point by laughing at him and/or levying accusations at him. Screw it, even if he is a troll he has a valid point, whether he knows it or not!
If you want to discredit him then do it with facts, don't do it by throwing out names like "troll" and then walking away as if you've settled some sort of cosmic score. Because you haven't. Me, I tend to think that there are just some things that we don't know yet, and might not ever know. I don't know (or particularly care) if gods, jesuses, devils, spooks, bunnies, or fairies are behind them, but I am at least honest enough to admit they can't be ruled out. Why isn't the rest of Slashdot like this?
Saying that the genome project was a "map to nowhere" is the same sort of neo-luddite crap we hear from people bashing pure research all the time. Whether it is measuring the age of the universe or decoding the human genome, the simple fact of the matter is that pure research is often done for just that... research.
There is no assured outcome any time you do pure research, but the knowledge continues to prove useful in many disperate fields. Anthropology, history, medicine... they have all had tangible results ALREADY from the work that was done.
Just because the initial "decoding" is done doesn't mean that the project is finished. Much like version 1.0 of software, there is much research and debugging to be done. Not really sure what the point of this article was... We don't have instant results on April 17, 2001??? Give it time.
I would have to say that explosives are the most abused technology in all of history.
Clearly. The immediate problems with the interpretation of this first, cursory glance at the genetic blueprint and it's dissemination are:
a) the lack of information (we'll understand more when we can explain the differences -- Celera loosely assembled 5 genomes, HGP [more completely] one),
b) the gap between the scientists and the data handlers (hopefully quickly filled by burgeoning bioinformaticians), and
c) (most notably in this case) the ignorance and chicken-littleism of the press.
The point is, we never would have gotten this far without a gene->protein hypothesis. Now is the time to understand the complexities of the system. There likely will not be a one description fits all remedy for the problem. There are plenty of problems with/exceptions to the rules. The only hard and fast rule is survival. And that one, as far as the general public is concerned, is a difficult hurdle to clear.
The Spectator
(I'm paraphrasing...mostly...exclamation points added for effect...mostly):
- DDT is good! It's really good! Rush even talked about it!
- Scripts and Goofs!
- The DNC, Unions, Dem. Davis, and Dem. Hillary are momos!
- Pres. Bush's Arsenic is good for you!
- People for the american way suck! (Psst And they like the Chinese)
- See Bush likes the environment (and he really hates China)! Jesse Helms in Mexico! 30% of blacks in Mississippi want the Confederate flag!
- The damned speak!
- Genes suck! Submit! Submit!
- Clinton took a trip and paid for it himself. The Clintons damaged the white house carpet. The Bush administration doesn't know how to keep track of resumes.
With all those stories I wonder how the genome story will be interpreted? Most likely with the unblinkingly analytical eye of science! Sheesh! I'd rather read a Salon writer compare sub-atomic particles to Survivor contestants for five pages than this pap.Deadly Green
(Mr. Shows "Scams and Flams" comes to mind)
Scripts and Goofs!
Buildings Romance
Poisoner-in-Chief Is Saving Lives
Creepy People for the American Way
The Greening of George W. Bush
(Quote:Amen, brother. Southern is the last endangered species, the last ethnic group it's still OK to taunt. If we must suffer to stem the tide of political correctness, then the slings and arrows of outrageous Yankees will have been worthily endured. And thank heavens we have a south-mouth in the White House!)
Winning in Hainan
Map to Nowhere
Clinton's Pay the Price
"I ain't got no flyin' shoes."
Forgive me if I'm wrong (and I invite criticism); my study of virii is not very in depth. However, viruses function in a very specific way.
To replicate, the virus injects its DNA into the host cell. The viral DNA then patches itself into the host DNA, so that you have the instructions for building the virii in the cell's chromosome. When the cell divides, this DNA data is copied, and you get multiple cells which create the virus.
However, the patching-in part of the viral replication process is very interesting. Here is an example strand of DNA:
GCGTGCCAGCAG
CGCACGGTCGTC
For the virus to patch itself in, it has to split the DNA where it can. So you'll end up with something like:
GCGTGCCA GCAG
CGCA CGGTCGTC
The single-sides parts of the DNA are commonly called sticky ends, where the viral DNA might patch it. However, the viral DNA has to patch in to a part that matches up with its own sticky ends; A (adenine) pairs only with T (thymine) and C (cytosine) pairs only with G (guanine). You can see that I've followed the pairing pattern consistently (pairs are listed vertically).
For viral DNA to fit into the above example, it must look like the following:
GCGT-GCCA |INTERIOR VIRAL DNA|-GCCA GCAG
CGCA CGGT-|INTERIOR VIRAL DNA| CGGT-CGTC
The viral DNA must have sticky ends that match. For cosmetic purposes, I hyphenated the example and spaced it so the pairs match up. The viral DNA has a left-end sticky end of CGGT, and a right-end sticky end of GCCA. These sequences match up with the corresponding sticky ends of the host's DNA, so the virus can 'patch' in and be replicated.
Now, to relate back to the story: if we have a genetic map, and we learn specifically where AIDS patches itself in, we can likely devise a blocking mechanism. Every three DNA base pairs is a codon for an amino acid, however, many of these codons are redundant; there are 8 (IIRC) different codons for leucine, an amino acid. So, if we could find which codons match up, we could possibly substitute a different codon at the viral injection point, but code for the same amino acid (so as to create the same protein).
Anyway, it's a leap, but most science is, at first.
[The following makes the rhetorical assumption that the parent is not a troll]
...and when we do figure it out, will you recant and become a card-carrying atheist?
...what happens is science does end up solving the problem, and your rational reason for believing in God instantly evaporates.
This kind of "God of the Gaps" argument for the existence of God will get you in trouble. Every time a representative of "The Church" makes the following claim:
"Science can't explain X, therefore it must be the work of a Supreme Being, which therefore must exist."
So what do you do then?
There are other reasons for believing in the mystical, and some of them are much easier to defend. Read about it. That goes for all o'y'all.
Obviously, it is. So where is this extra information located? It is obvious that there must be some other mechanism at work. I would posit that the mechanism is supernatural.
There really is no other explanation. The Church has known this for many thousands of years, and now the scientists are realising it too. The missing information must be supplied by the Holy Spirit. When a man impregnates a woman, the Holy Spirit breathes life into the resulting embryo. At least, this is what we were told in school. In actual fact, it breathes information in, and gives it a soul.
I know I will be labeled a troll for this, and am saddened. But really, the Church has known this for thousands of years, and now we are being proved correct.
Looking for the good news? Here it is:
"Nonetheless, Celera's message is not likely to comfort investors. Gene therapy holds out less promise as a result of this new understanding."
That's right, the Rockville MD based company that is busy literally patenting our asses has just discovered that it doesn't know what it's patenting. The whole model of patenting a gene that codes for a protein has fallen apart, since with 10 times as many proteins as genes, we don't really know what genes do anymore.
Hallelujiah.