Posted by
timothy
on from the spin-the-wheel-of-fortune-dial dept.
GuyFromAccounting writes: "Tomorrow's Economist has a new article
describing research that shows that the number of human genes is more than twice
the estimate made a few months ago.The
article describes why it is so hard to estimate the number of genes."
If you see a pattern in something, there are two possibilities:
The pattern is a result of your measurement technique.
The pattern is a result of something external. (i.e., something you actually care about.)
That is to say, science articles (and especially Slashdot articles) tend to exaggerate either the scope of someone's research or their confidence in it. Actual articles in science journals tend to be very conservative in their claims, often mind-numbingly so. Conjectures and guesses get inflated in popular science articles to the level of "scientific truth", when in fact actual scientists in the field aren't entirely convinced.
It is quite possible that science articles seem so contradictory because that is the way they are written. Contradiction and conflict are dramatic and interesting. I think science writers emphasize this to make their articles more appealing. This skewed presentation may be the cause of what you are observing.
That isn't to say that the body of scientific knowledge never changes. It usually changes in a much more laid-back and boring way than portrayed in the media, however.
God is a prima donna programmer
by
Zombie
·
· Score: 5
This makes me think of the item on "prima donna" programmers the other day. If we for a second assume that the preposterous be true, and there is a God and he created man, then
he's a spaghetti coder
he doesn't document any of his code
he supports the "embrace and extend" method of evolving functionality
he's clearly into this whole job security through obscurity thing
lots of people die because we can't figure out his code, so he doesn't give a rodent's behind about us
Actually... it looks like God works for Bill Gates!
Re:God is a prima donna programmer
by
PD
·
· Score: 2
That may be true, but he wrote everything in BrainFuck
Re:God is a prima donna programmer
by
istartedi
·
· Score: 2
Yes, but He didn't give us the source. Therefore, all of you in the "closed source is immoral" camp must be either atheists or blasphemers. I know Stallman is an atheist; I guess that makes a lot more sense now.
-- For all intensive purposes, "whom" is no longer a word. That begs the question, "who cares"?
Re:God is a prima donna programmer
by
joto
·
· Score: 5
Ah, no. But what you see, is only the binary. It's only after it is compiled with gcc (genetic code compiler) that it seems messy. It's like viewing a binary in texteditor. As I remember, God used to be especially picky on intendation and variable naming, and listed preconditions and postconditions for all his functions, as well as invariants for all of his datastructures.
Re:'Genes' vs. 'Instructions'
by
/dev/kev
·
· Score: 2
Nature isn't that wasteful, and it wouldn't carry around 90% of the DNA for no good reason.
That's right, Darwin's theory of evolution says that. The thing is, just being there might be the only reason needed. Padding is a perfectly valid possible use/function. When these people say "useless", they mean "no direct functionality, unlike these other interesting parts over here."
If they don't know what it means, they should just say so and keep working at it.
They do. They say "We don't know what this means, but here's our best guess...". The problem is that all too often the media takes the best guess and reports it with more certainty than it deserves.
IANAG (I am not a geneticist), but one argument I've heard about the so-called "meaningless drivel" in the DNA is that the current estimates for the distribution of gene sizes is heavily biased towards small genes. Apparently, the cost to sequence a gene scales with the size of the gene, and in the interest of actually finishing a sequencing project, geneticists have favored the smaller, and therefore cheaper genes. The estimates for amount of "meaningless drivel" simply take the estimate for average gene size multiplied by the estimate for the number of genes, and find that this falls short of the total number of base pairs. The problem, then, is that the average size of genes is severely underestimated.
-- ***
Work like a king, command like a slave, create like a dog.
Aint it cool? According to the article, Humans are actually coded OO style:-)
"For example, 60% of "zinc finger" genes (whose protein-products help to regulate the expression of other genes), are located on chromosome 19. It looks as though they have evolved by repeated duplication from a single "grandmother" gene, followed by specialisation to do slightly different jobs. Protein-kinase genes, whose products are involved in intracellular signalling, are similarly concentrated on chromosome 1. The researchers tripled the number of protein-kinase genes known from this chromosome. They also found hints that genes whose protein-products work together in a cell have sometimes ended up as neighbours on a chromosome. That might simplify the co-ordination of their expression."
Reminds me of GA-designed circuits
by
Hard_Code
·
· Score: 2
This reminds me of the GA-designed circuits story...in which a researched ran an GA program to design a circuit for a specific purpose, and after some evolution got a circuit which did the job correctly. It was composed of a main circuit which appeared to do all the work, and a totally seperate, isolated loop disconnected from the rest of the circuit, which for all intents and purposes appeared to do nothing and be completely useless. So he removed this "useless" loop, and suddenly the circuit as a whole stopped functioning. This seemingly useless piece was actually manifesting some spooky effect on the rest of the circuit. It just goes to show that evolution cares only about results, not process. Which explains the amount of "drivel" we have in our genome, and the gross energetic inefficiency of most life. Sometimes I wonder what would happen if we engineered our own organism from scratch, and left out all of the cruft. It would probably be super-efficient, except it may not be so robust against "unenvisioned" problems (ambiant radiation?). That "junk" must serve some buffering purpose (think of algorithms which employ randomness - they're not as omptimal under the best circumstances, but they completely avoid pathological cases).
Re:'Genes' vs. 'Instructions'
by
dodobh
·
· Score: 2
No, the entire genome is the OS. The genes are userland programs, while the rest is kernel code, and no-ops.
Currently we are hacking at the Userland programs, then we will become kernel hackers;).
No. Sometime next Tuesday, Cowboy Neil convinces me to loan Rob mine. Well, I say next Tuesday loosly, because, well, it hasn't happened yet. I could change my mind, and not drive over there until Wed, or I could get sloshed on Bass Ale and not do it until Thurs.. Or could I? I mean, it's happening, well, that Rob came back from the future with me on Tues, I mean, that is what I told me, so..
But.. But..
I gotta go.. My head hurts.
-- .sig: Now legally binding!
Re:open source and cooperate?
by
lovebyte
·
· Score: 2
Maybe if they open sourced their efforts and cooperated
1- Most databases are publicly available.
2- Many bioinformatics groups DO cooperate
Unfortunately, everybody has a different view of what is a gene and how to find them.
--
I'll do it for cheesy poofs.
Re:'Genes' vs. 'Instructions'
by
krmt
·
· Score: 2
Not at all. As I stated in my previous post, DNA is not code. It's an analogy, but the simple fact is that proteins are not hardware, but rather they are a part of the whole system.
If you really want to look at a biological system as a computer program, it's a better idea to think of the proteins as part of the program, rather than the thing the program runs on.
The DNA contains instructions, but in no real linear order. That's one (of about a billion) reason why deciphering the human genome is so difficult. Granted, source code isn't necessarily linear either, but there is no entry point for the "program" of a living system. Even birth is not a starting point because there's already a bunch of stuff happening even before the cell is implanted (all life from life). While this can be likened to a compiler, it sheds light on the idea that DNA isn't so much a program itself but part of this gigantic system which includes a lot of other "programs." Just as a kernel can not really do much on its own, neither can DNA.
I'm not saying DNA is not like code, indeed the fact that we use the term "genetic code" at all indicates that there is some degree of similarity. But to consider DNA to be the entire thing is completely absurd. You can understand a program by looking at the source, but that's like saying you can understand a gene by looking at its protein sequence and such (which isn't completely true, but you can deduce a fair amount from it). You are looking at an entire system here, not just a single "program". It's actually very similar to a huge UNIX system, with a ton of small programs (proteins) that each do a few specific tasks. You can't understand UNIX by looking at one program and ignoring the rest. You also can't understand UNIX by looking at all the source code together (especially if you don't understand most of it, as we don't) because you don't know what programs are running at what time. Just because you have the source code to apache doesn't mean it's running right now. Same thing with proteins, you don't know if they're working right now. That's a major major problem, and it's one that neither the UNIX source nor the DNA has an answer to.
A gene has been very specifically defined. It goes as follows: "a gene is a unit of heredity." That's all there is to it. All the rest is a function of that. DNA is a physical manifestation of this idea, and the different genes must all perform different functions, but all are units of heredity that are passed on to subsequent generations.
"I may not have morals, but I have standards."
--
"I may not have morals, but I have standards."
Re:'Genes' vs. 'Instructions'
by
krmt
·
· Score: 2
Ok, fair enough, perhaps saying that they can't do anything by themselves is an exageration, but they can't produce life by themselves, which is really the key. You can't consider RNA and DNA on their own because they belong in a complete system. Even viruses require a protein coat, and they're in between in the life/not-life category. To understand anything living, you need the proteins too.
"I may not have morals, but I have standards."
--
"I may not have morals, but I have standards."
Re:'Genes' vs. 'Instructions'
by
krmt
·
· Score: 2
Actually, the synthesis of this information is one of the branches of the research I am doing right now... and if you'd like to dicuss it further, I wouldn't mind
Cool! What kind of stuff? I'm always game:-)
"I may not have morals, but I have standards."
--
"I may not have morals, but I have standards."
Re:'Genes' vs. 'Instructions'
by
krmt
·
· Score: 5
blockquoth eraserbones
And yet, the DNA spans between genes are generally referred to as 'useless' or, in this case, 'meaningless drivel.' Am I missing something, or is this exactly where the good stuff is?
Not necessarily. You're caught in they typical/. trap of thinking of DNA as computer code. Granted, code is probably the best analogy we have, but it's still an analogy at best. You're very correct that living systems can operate by branching, looping, etc. just like programs. However, your mistake lies in looking too hard at the DNA and not hard enough at the whole system.
The fact is that DNA itself is pretty useless. It can't do anything without proteins, and it's the proteins that are actually acting on each other, on the DNA, and on the RNA. That said, it's probably the proteins that allow these functions in terms of things like splicing out introns (alternate splicing is a form of branching) and DNA replication via DNA polymerase and other helpers (a form of recursion).
While I personally don't believe that the intervening DNA sequences are complete garbage, I don't think they hold the processes you're looking for as much. I agree with the idea that they provide a lot of raw genetic material for evolution, and I also think they play a role in gene regulation by chromatin bundling and such.
However, the idea of DNA as a program is only a small part of the picture, and in reality even when we have the genome and the proteome, we're still going to have to figure out how everything works together. A living system is big and complex, with tons of parts we don't understand yet. It's going to be a fun time figuring it out.
"I may not have morals, but I have standards."
--
"I may not have morals, but I have standards."
Re:'Genes' vs. 'Instructions'
by
Dirtside
·
· Score: 2
No, no, the introns are *source comments*:
ACGTCATCTGAGCGTCGCGGCAGTAGTCTGCGTATGCTGAGTCGAGC
/* Pinky finger - this should be the right size to fit comfortably into the hole in a CD */
So you're saying no science should be published until it satisfies a need you see as useful? This is why there's so much bad science: people insisting that there's an industrial need for the product of all research. How do you think arseholse like Monsanto got the idea to patent genes and screw us all out of the benefits of our own genetic heritage? Through narrow-minded thinking like yours.
Anyway, now I'm done with you, I'd like to point out that identifying a 'use' for every gene is NOT the best goal of genomics. There's a great line of thinking which says that mobile gene fragments may have transported modular portions of genes around the genome all the way from intracellularly to inter specially. If common structures can be found throughout the genome, particularly in functional gene exons, it might prove that the path of evolution is alot more interesting and perhaps short than we thought.
Patterns of punctuated equilibrium dominate modern theories of special evolution. For instance, the mammalian radiation after the death of the dinosaurs. We went from a few small rodent-types to a gigantic variety of fauna in a few million years. Identifying links between these species on modular gene and non-encoding DNA fragments can tell us about geography, nationality, immunity, evolution - damn near everything.
The genome is a fascinating place. Don't get trapped into thinking that the genes are the only interesting part. Maybe the uses above aren't the single most important right now, but the more we know about the genome, the more ways we have to exploit it and stop the onslaught of terrible genetic disease. Many great inventions occur by accident, so you should never limit science to what you see as strictly 'useful.'
Re:'Genes' vs. 'Instructions'
by
localman
·
· Score: 2
I'm glad to hear there are others who have realized this. I agree wholeheartedly. It's bugged me quite a bit that the folks doing this research feel confident that more than 90% of our genetic information is meaningless.
What would people think if someone claimed to finally decode the mysterious "Russian" language, and announced that it was 90% meaningless?
Nature isn't that wasteful, and it wouldn't carry around 90% of the DNA for no good reason. Didn't some doctors used to think that the heart was a useless organ? Even the tonsils are somewhat useful. The appendix seems to be one of the few examples of wastefulness. So maybe if I heard that less than 10% of DNA was meaningless I could believe it.
Fact is, we barely know anything about what our DNA means. I'd bet there's a lot more to it than the straight protien mappings that we know about now. I am very excited to see the developments over my lifetime in figuring it out. But when I hear these reports, I wonder if the folks doing it are even qualified. If they don't know what it means, they should just say so and keep working at it.
Back at TAMS (the Texas Academy of Math and Science), clubs sprung up like wildflowers. Most of them were quite silly. But, the worst part, was that they would spam you - in everyone's box, they'd put all sorts of unwanted advertisements. So, we decided to make a counter-club. We formed the Time Traveler's Club (TTC). We put up big signs with our slogan, "I'll See You Yesterday!". We passed out notes announcing that our next meeting would be last Tuesday. We passed out notes thanking everyone for the large turnout at our last meeting on the following friday.;) Ah, that place was fun... that was almost as fun as the "Nobody For President" campaign we ran...
-= rei =-
-- "Well, then fire it up and show me what this..." (sigh)... "coccoon can do."
A perfect example of this is cloning. When Dolly was cloned, the media went all-out in seing cloning as "Its here!!" Then, when problems started showing up, people started on the "cloning is horrible, everyone gets defects" bandwagon. When, in reality, neither were true.
The technique used in creating dolly was just awful. The scientist who worked on it has become a cloning opponent, largely due to seing his failures. On the contrary, other cloning researchers, like the ones that did the honolulu experiment with mice, have become its biggest proponents.
What was the dolly experiment like? Well, first off, they chose sheep because there is a very long period of time from when the egg is fertilized to when it divides for the first time; this length of time was assumed to (and likely does) help the odds of the cell thinking things were normal. However, his techniques were awful. After denucleating the ovum, it would go into a dormant state. So, he had to get the new nucleus to be in a dormant state, too. He did this by starving the cell for hours until it almost died. Then, instead of transplanting the nucleus directly, he applied a powerful electric shock through the solution the two cells were in, which usually caused them to merge, and act like a just fertilized egg. Now, I'm sure everyone reading this is just going, "this is going to cause serious problems". And, that it does. Dolly was a lucky sheep. Most of the embryos weren't near as lucky - that shock does a lot of damage to the cell (and starving the nucleus until it shuts down is bad too).
The honolulu experiments, for contrast, used mice. Mice are an even harder subject to deal with, because they have notoriously fast divisions in fertilized egg cells. But, they used them anyways, because they were not only convinced they could clone them, but they wanted to see results several cloned generations down the line. The technique used there involved, like before, denucleating the egg cell - but, doing this *right before* having the new nucleus implanted. To get a dormant nucleus, he took cells that were always (or almost always) dormant, such as certain nerve cells. He extracted the nucleus, and implanted it immediately into the egg cell. Then, he put the new egg cell inside a solution which prevents it from forming a polar body (and throwing away half the genes), so that it would think it was now fertilized. It then began to divide. They had a level of success that looks like, once the technique is perfected, will approach normal external fertilization techniques. Signs of premature aging didn't occur until 5 generations of clones - this due to the fact that genes slowly change over time, and usually for the worse; we basicly extended the mouse's lifespan beyond what its DNA was designed to handle (this could be fixed by making a "DNA backup" from the original mouse, and then reconstructing that DNA each time you want a clone - an important reason for fast DNA sequencers).
The media locked onto the first story. When problems started to arise, they completely switched gears, and made it look like all cloning is dangerous. Bad media! No cookie!:)
-= rei =-
P.S.: BTW, I have yet to see a project where they actually transfer the mitochondria - that's over 1% of most animals' DNA.
-- "Well, then fire it up and show me what this..." (sigh)... "coccoon can do."
The reason is that the genes don't exactly have a marker at each end to delineate them. The genes are to some extent a matter of definition. They're different lengths, they can sometimes be found in different places, and sometimes 'two genes' do exactly the same thing even though they read quite differently.
Some bacteria have been found with two or three sets of genes sort of ontop of each other- starting with different offsets. Its a bit like code that you can jump to at 0x2000 or 0x2001 and it does different, but useful things!
Anyway reverse engineering this lot will take a while...
Month-old stories used to get posted, but no we are getting links to stories that aren't even available yet! Maybe Taco finally got his time machine that was backordered at ThinkGeek.
Re:'Genes' vs. 'Instructions'
by
update()
·
· Score: 2
To add a bit to krmt's excellent points:
If you insist on thinking of DNA as C code, think of the coding sequences as printf's and such and the intervening segments as the rest of the code -- flow control and whitespace. (We're talking about C, not Python..) Humans have a huge amount of whitespace, most of which probably isn't important at all. But it's very difficult to distinguish regulatory elements ("if" statements) from junk (whitespace).
On the whole, you're right, though. The raw number of genes is more media-friendly than scientifically important. Unless your company has a business plan based on patenting genes.
Unsettling MOTD at my ISP.
Nothing is remotely firm yet...
by
update()
·
· Score: 4
The article (and the writeup here) makes it sounds like one presumably very accurate estimate has been supplanted by a very different, presumably very different, estimate. The reality is that identifying genes in raw sequence is very much a work in progress. At the annual Genome Sequencing meeting at Cold Spring Harbor in May, a bunch of groups presented different methods that resulted in widely divergent numbers. Everyone's numbers were increasing over the estimates of last year, though.
It'll sort itself out over the next couple of years as the sequence gets better assembled, more non-human sequence is available for comparison and the groups adopt one another's good ideas. In the meantime, it looks like a good PR person at Ohio State managed to make their findings seem more revolutionary than they are.
By the way, if you want to bet on the number, see the GeneSweep page. (Note that bets must be placed in person!) I put my $5 on 44,000 and change.
The plain old numbers game doesn't work with the genome - we have 3 billion base pairs but frogs have 9 billion. We've got 46 chromasomes but dogs have 76.
The most telling figure is that we're 99%++ genetically identical to chimpanzees and yet we can't interbreed (not that I've tried...).
What really matters on a genomic level is the interplay between genes during crucial times of development, not just the functions of individual genes (our ribosomes are nearly identical to bacterial ones) - many of which were selected by evolutionary pressures meaning that once a problem had been solved, for example how to copy DNA, that gene was 'set in stone'. After all, who wants to keep solving a problem once you've found a solution?
Back to the point on crucial times of development - could you imagine what the result would be if the gene(s) controlling synapse differentiation in the fetus stayed active for an extra hour? day? week? - that, ladies and gentlemen, may be all that is separating us from chimpanzees. Just a simple interplay of genes, subtly disturbed by a simple mutation, perhaps lengthening a crucial phase in brain development.
So the numbers game is just irrelevant.
-Nano.
It's psychology. While most scientists tend to regard the first few studies on a topic as little more than a theory until its been confirmed by a few other people, the Media and general public tend to take these as absolute answers. "Well, then" they say "Thats taken care of". If the next study indicates the first one is wrong, then somethings changed. Thats news. If it simply confirms that yes, we are little more complex than your average nematode, thats not news. If suddenly we have way more, then the number of genes in the human genome has changed(well not really but you understand what I mean) thats news and shows up in the popular media. Slashdot works the same way. Taco and company aren't going to post 14 stories confirming the number of genes - it's not new, and (to most people) not exciting. But if it challenges current beliefs, its exciting and gets posted.
As for a health skepticism about science, you should be skeptical about science. Skepticism is (or should be) an integral part of science. Nothing should be taken for granted, nothing should be accepted as true until a good number of people have had a chance to kick it around in every way they can think of without finding a problem with the theory/study/whatever.
And when evidence does surface proving that the last theory was wrong, a new one should be created to fit the data and then that should be put under the microscope for flaws. This isn't religion. You don't take things on faith. Everything should be questioned and tested before acceptance.
-- Why?
URL of the article in Genome Biology
by
mcockerill
·
· Score: 4
The research article is available from the Genome Biology web site here.
Number of possible genes
by
Socrates0
·
· Score: 2
Because of the way that Celera (and, AFAIK the HGP) identify genes, I feel that we are going to be seeing this soft of announcement increasingly often. Celera has been using imprinted cells as a base for their sequencing. Imprinted cell have many genes hidden or 'turned off' by interceptor protiens. What needs to be done, but could take many years, and more bio-chemistry than we have now, is to 'walk' the genome. Take the base pairs 3n at a time, and sequence, ala Folding@Home. Actual protien sequences overlap, in a sort of bio-chemical data compression. If you offset the "start here" tag by one base pair you don't always end up with a garbage protien. This is why Celera's claim of such large quantitys of 'junk' DNA was scoffed at by many in the scienfific community.
But I'm rambling. Time to go home.
Re:Number of possible genes
by
Socrates0
·
· Score: 2
1) I was under the impression that Celera's technique involved finding a protien and following it back to the gene, then sequencing. It was supposed to "Only sequence 'Real Genes (tm)' and not all that filler".
2) 3n at a time. As in, n codons at a time (DNA is meaningless other wise). I was trying to suggest that genes have both a location and an offset, so we should sequence a given series of codons 3 times, with different offsets. Yes, it's more work, but certain features of viri suggest that the same stretch of DNA can do multiple protiens.
3) Actually, more than 90% of DNA is non coding in any given cell. All cells will sequence the Citric acid cycle for instance, but only a few will sequence seratonin. You MUST sequence (and compile the protien) for the whole genome. Then, look at another type of cell to see what you missed. Mishapen protien fragments that seem to be "junk" may shape polymorphic protiens as they fold. A sequence lacking its acton may suddenly have one when other "junk" is removed in an intermediate step. Just because a gene is nonexpressed or seemingly non-expressable does not mean it has no function.
open source and cooperate?
by
NaturePhotog
·
· Score: 2
The result is a lot of data, but those data are scattered in numerous databases that are organised and maintained in diverse ways by various research teams.
Maybe if they open sourced their efforts and cooperated, (a) they'd spend less time potentially looking at the same things and inventing the same processes (b) somebody would get a better sense of how much has been done and how much is left to go by looking at all the data.
I know, I know...there wouldn't be any patents and piles of money if they did it that way, hence they have no motivation to...
And of course, then they'd then spend their time fighting off German lawsuits for naming their sequencing software KGene or some such...
Gene Anderson
Did Venter and Collins celebrate to early???
by
andres32a
·
· Score: 4
When Venter and Collins announced some months ago their "success" at sequencing the human genome and that according to their conclusions humans did not have much more genes that worms this did seem kind of odd to me.
If this new paper published by "Genome Biology" is right it could ridicule all past celebration... i mean... if it turns out that Venter only mapped 30 000 genes but there were actually 60 000 genes in humans, wouldnt that mean that were not even half way to a complete gene map for humans??
One would hope that we don't hear the same article posted in 6 months, lest a Moore's Rule of Genetics come into play.
Kudos, Scientists, for doubling your workload.
(I personally would say more but I didn't take Genetics due to the 8am lab.)
Isn't that what the census is for?
by
Anonymous+DWord
·
· Score: 4
Personally I don't know any human Genes, although my friend had a hamster named Gene. It died, unfortunately. You'd think it would be harder to count the non-human Genes, but I guess that's far in the future. Didn't they count the mouse Genes a while back though?
-- "If he thinks he can hide and run from the United States and our allies, he's sorely mistaken." Bush on bin Laden
Re:'Genes' vs. 'Instructions'
by
eraserbones
·
· Score: 2
I agree with everything you've said, but I'd still like to put a bit more muscle behind the 'DNA as code' analogy. Ignoring the role that the ambient cellular environment plays would be foolhardy, true. But so would ignoring the role that hardware plays in the interpretation of software. The thing that makes the DNA the interesting part is that cells are the same from organism to organism, whereas DNA varies. Similarly with hardware and software.
It feels to me like your criticism could be leveled at a desire to understand a program by looking at the source code. "the actual branching, jumping and iteration happens in the ALU, not in the code, so the code is the wrong place to look," and so on.
'Genes' vs. 'Instructions'
by
eraserbones
·
· Score: 4
Genes, I'll grant you, are the exciting bits of a chromosome, because they (generally) correspond to proteins that can be identified and detected. But, I'm not entirely clear on why they are the primary focus of genetic research.
We all seem pretty comfortable discussing DNA as though it were computer code, so let's follow that metaphor a little further. If I point at a big mess of C code (say, a console app) and ask 'What does this code do?' an amateur might be tempted to scan it for printfs, puts's, and other 'output signifiers.' But really, if that's all you look at, you don't have a clue what the actual funtion of the code is. All those boring scanf's, if/thens and operators are really important.
My rudimentary education in genetics has me convinced that DNA in a living cell has the ability (like C code) to switch, jump, branch, and (most importantly) operate recursively on its own resultant proteins. And yet, the DNA spans between genes are generally referred to as 'useless' or, in this case, 'meaningless drivel.' Am I missing something, or is this exactly where the good stuff is?
And, viewed from this angle, isn't counting genes as pointless as counting KLOCs?
Slashdot Mirror
That is to say, science articles (and especially Slashdot articles) tend to exaggerate either the scope of someone's research or their confidence in it. Actual articles in science journals tend to be very conservative in their claims, often mind-numbingly so. Conjectures and guesses get inflated in popular science articles to the level of "scientific truth", when in fact actual scientists in the field aren't entirely convinced.
It is quite possible that science articles seem so contradictory because that is the way they are written. Contradiction and conflict are dramatic and interesting. I think science writers emphasize this to make their articles more appealing. This skewed presentation may be the cause of what you are observing.
That isn't to say that the body of scientific knowledge never changes. It usually changes in a much more laid-back and boring way than portrayed in the media, however.
Actually... it looks like God works for Bill Gates!
Nature isn't that wasteful, and it wouldn't carry around 90% of the DNA for no good reason.
That's right, Darwin's theory of evolution says that. The thing is, just being there might be the only reason needed. Padding is a perfectly valid possible use/function. When these people say "useless", they mean "no direct functionality, unlike these other interesting parts over here."
If they don't know what it means, they should just say so and keep working at it.
They do. They say "We don't know what this means, but here's our best guess...". The problem is that all too often the media takes the best guess and reports it with more certainty than it deserves.
Quidquid latine dictum sit, altum viditur.
*** Work like a king, command like a slave, create like a dog.
You mean you just figured this out?
To within half a percent, pi seconds is a nanocentury. -- Tom Duff
Aint it cool? According to the article, Humans are actually coded OO style :-)
"For example, 60% of "zinc finger" genes (whose protein-products help to regulate the expression of other genes), are located on chromosome 19. It looks as though they have evolved by repeated duplication from a single "grandmother" gene, followed by specialisation to do slightly different jobs. Protein-kinase genes, whose products are involved in intracellular signalling, are similarly concentrated on chromosome 1. The researchers tripled the number of protein-kinase genes known from this chromosome. They also found hints that genes whose protein-products work together in a cell have sometimes ended up as neighbours on a chromosome. That might simplify the co-ordination of their expression."
This reminds me of the GA-designed circuits story...in which a researched ran an GA program to design a circuit for a specific purpose, and after some evolution got a circuit which did the job correctly. It was composed of a main circuit which appeared to do all the work, and a totally seperate, isolated loop disconnected from the rest of the circuit, which for all intents and purposes appeared to do nothing and be completely useless. So he removed this "useless" loop, and suddenly the circuit as a whole stopped functioning. This seemingly useless piece was actually manifesting some spooky effect on the rest of the circuit. It just goes to show that evolution cares only about results, not process. Which explains the amount of "drivel" we have in our genome, and the gross energetic inefficiency of most life. Sometimes I wonder what would happen if we engineered our own organism from scratch, and left out all of the cruft. It would probably be super-efficient, except it may not be so robust against "unenvisioned" problems (ambiant radiation?). That "junk" must serve some buffering purpose (think of algorithms which employ randomness - they're not as omptimal under the best circumstances, but they completely avoid pathological cases).
It's 10 PM. Do you know if you're un-American?
No, the entire genome is the OS. The genes are userland programs, while the rest is kernel code, and no-ops. ;).
Currently we are hacking at the Userland programs, then we will become kernel hackers
I can throw myself at the ground, and miss.
No. Sometime next Tuesday, Cowboy Neil convinces me to loan Rob mine. Well, I say next Tuesday loosly, because, well, it hasn't happened yet. I could change my mind, and not drive over there until Wed, or I could get sloshed on Bass Ale and not do it until Thurs.. Or could I? I mean, it's happening, well, that Rob came back from the future with me on Tues, I mean, that is what I told me, so..
But.. But..
I gotta go.. My head hurts.
.sig: Now legally binding!
Maybe if they open sourced their efforts and cooperated
1- Most databases are publicly available.
2- Many bioinformatics groups DO cooperate
Unfortunately, everybody has a different view of what is a gene and how to find them.
I'll do it for cheesy poofs.
Not at all. As I stated in my previous post, DNA is not code. It's an analogy, but the simple fact is that proteins are not hardware, but rather they are a part of the whole system.
If you really want to look at a biological system as a computer program, it's a better idea to think of the proteins as part of the program, rather than the thing the program runs on.
The DNA contains instructions, but in no real linear order. That's one (of about a billion) reason why deciphering the human genome is so difficult. Granted, source code isn't necessarily linear either, but there is no entry point for the "program" of a living system. Even birth is not a starting point because there's already a bunch of stuff happening even before the cell is implanted (all life from life). While this can be likened to a compiler, it sheds light on the idea that DNA isn't so much a program itself but part of this gigantic system which includes a lot of other "programs." Just as a kernel can not really do much on its own, neither can DNA.
I'm not saying DNA is not like code, indeed the fact that we use the term "genetic code" at all indicates that there is some degree of similarity. But to consider DNA to be the entire thing is completely absurd. You can understand a program by looking at the source, but that's like saying you can understand a gene by looking at its protein sequence and such (which isn't completely true, but you can deduce a fair amount from it). You are looking at an entire system here, not just a single "program". It's actually very similar to a huge UNIX system, with a ton of small programs (proteins) that each do a few specific tasks. You can't understand UNIX by looking at one program and ignoring the rest. You also can't understand UNIX by looking at all the source code together (especially if you don't understand most of it, as we don't) because you don't know what programs are running at what time. Just because you have the source code to apache doesn't mean it's running right now. Same thing with proteins, you don't know if they're working right now. That's a major major problem, and it's one that neither the UNIX source nor the DNA has an answer to.
"I may not have morals, but I have standards."
"I may not have morals, but I have standards."
A gene has been very specifically defined. It goes as follows: "a gene is a unit of heredity." That's all there is to it. All the rest is a function of that. DNA is a physical manifestation of this idea, and the different genes must all perform different functions, but all are units of heredity that are passed on to subsequent generations.
"I may not have morals, but I have standards."
"I may not have morals, but I have standards."
Ok, fair enough, perhaps saying that they can't do anything by themselves is an exageration, but they can't produce life by themselves, which is really the key. You can't consider RNA and DNA on their own because they belong in a complete system. Even viruses require a protein coat, and they're in between in the life/not-life category. To understand anything living, you need the proteins too.
"I may not have morals, but I have standards."
"I may not have morals, but I have standards."
Actually, the synthesis of this information is one of the branches of the research I am doing right now... and if you'd like to dicuss it further, I wouldn't mind
:-)
Cool! What kind of stuff? I'm always game
"I may not have morals, but I have standards."
"I may not have morals, but I have standards."
The fact is that DNA itself is pretty useless. It can't do anything without proteins, and it's the proteins that are actually acting on each other, on the DNA, and on the RNA. That said, it's probably the proteins that allow these functions in terms of things like splicing out introns (alternate splicing is a form of branching) and DNA replication via DNA polymerase and other helpers (a form of recursion).
While I personally don't believe that the intervening DNA sequences are complete garbage, I don't think they hold the processes you're looking for as much. I agree with the idea that they provide a lot of raw genetic material for evolution, and I also think they play a role in gene regulation by chromatin bundling and such.
However, the idea of DNA as a program is only a small part of the picture, and in reality even when we have the genome and the proteome, we're still going to have to figure out how everything works together. A living system is big and complex, with tons of parts we don't understand yet. It's going to be a fun time figuring it out.
"I may not have morals, but I have standards."
"I may not have morals, but I have standards."
No, no, the introns are *source comments*:
T GT AGCTAGTCAGTCGTACGTAGTCG
ACGTCATCTGAGCGTCGCGGCAGTAGTCTGCGTATGCTGAGTCGAGC
/* Pinky finger - this should be the right size to fit comfortably into the hole in a CD */
GTCGTGTCAGTTGCATGCGTAGTCATCTGACGTAGTCTGACTGATGC
/* Appendix - I don't remember what this is for, but let's leave it in anyway */
And so on.
"Destroy science and religion. Science would re-emerge exactly the same; but not religion." - Penn Jillette, paraphrased
So you're saying no science should be published until it satisfies a need you see as useful? This is why there's so much bad science: people insisting that there's an industrial need for the product of all research. How do you think arseholse like Monsanto got the idea to patent genes and screw us all out of the benefits of our own genetic heritage? Through narrow-minded thinking like yours.
Anyway, now I'm done with you, I'd like to point out that identifying a 'use' for every gene is NOT the best goal of genomics. There's a great line of thinking which says that mobile gene fragments may have transported modular portions of genes around the genome all the way from intracellularly to inter specially. If common structures can be found throughout the genome, particularly in functional gene exons, it might prove that the path of evolution is alot more interesting and perhaps short than we thought.
Patterns of punctuated equilibrium dominate modern theories of special evolution. For instance, the mammalian radiation after the death of the dinosaurs. We went from a few small rodent-types to a gigantic variety of fauna in a few million years. Identifying links between these species on modular gene and non-encoding DNA fragments can tell us about geography, nationality, immunity, evolution - damn near everything.
The genome is a fascinating place. Don't get trapped into thinking that the genes are the only interesting part. Maybe the uses above aren't the single most important right now, but the more we know about the genome, the more ways we have to exploit it and stop the onslaught of terrible genetic disease. Many great inventions occur by accident, so you should never limit science to what you see as strictly 'useful.'
toeslikefingers.com - because
"Uh...yeah...turns out there are plenty more of them. If you want to see the data, though, you need to pay us royalties!"
Got Rhinos?
What would people think if someone claimed to finally decode the mysterious "Russian" language, and announced that it was 90% meaningless?
Nature isn't that wasteful, and it wouldn't carry around 90% of the DNA for no good reason. Didn't some doctors used to think that the heart was a useless organ? Even the tonsils are somewhat useful. The appendix seems to be one of the few examples of wastefulness. So maybe if I heard that less than 10% of DNA was meaningless I could believe it.
Fact is, we barely know anything about what our DNA means. I'd bet there's a lot more to it than the straight protien mappings that we know about now. I am very excited to see the developments over my lifetime in figuring it out. But when I hear these reports, I wonder if the folks doing it are even qualified. If they don't know what it means, they should just say so and keep working at it.
Back at TAMS (the Texas Academy of Math and Science), clubs sprung up like wildflowers. Most of them were quite silly. But, the worst part, was that they would spam you - in everyone's box, they'd put all sorts of unwanted advertisements. So, we decided to make a counter-club. We formed the Time Traveler's Club (TTC). We put up big signs with our slogan, "I'll See You Yesterday!". We passed out notes announcing that our next meeting would be last Tuesday. We passed out notes thanking everyone for the large turnout at our last meeting on the following friday. ;) Ah, that place was fun... that was almost as fun as the "Nobody For President" campaign we ran...
-= rei =-
"Well, then fire it up and show me what this..." (sigh)
A perfect example of this is cloning. When Dolly was cloned, the media went all-out in seing cloning as "Its here!!" Then, when problems started showing up, people started on the "cloning is horrible, everyone gets defects" bandwagon. When, in reality, neither were true.
:)
The technique used in creating dolly was just awful. The scientist who worked on it has become a cloning opponent, largely due to seing his failures. On the contrary, other cloning researchers, like the ones that did the honolulu experiment with mice, have become its biggest proponents.
What was the dolly experiment like? Well, first off, they chose sheep because there is a very long period of time from when the egg is fertilized to when it divides for the first time; this length of time was assumed to (and likely does) help the odds of the cell thinking things were normal. However, his techniques were awful. After denucleating the ovum, it would go into a dormant state. So, he had to get the new nucleus to be in a dormant state, too. He did this by starving the cell for hours until it almost died. Then, instead of transplanting the nucleus directly, he applied a powerful electric shock through the solution the two cells were in, which usually caused them to merge, and act like a just fertilized egg. Now, I'm sure everyone reading this is just going, "this is going to cause serious problems". And, that it does. Dolly was a lucky sheep. Most of the embryos weren't near as lucky - that shock does a lot of damage to the cell (and starving the nucleus until it shuts down is bad too).
The honolulu experiments, for contrast, used mice. Mice are an even harder subject to deal with, because they have notoriously fast divisions in fertilized egg cells. But, they used them anyways, because they were not only convinced they could clone them, but they wanted to see results several cloned generations down the line. The technique used there involved, like before, denucleating the egg cell - but, doing this *right before* having the new nucleus implanted. To get a dormant nucleus, he took cells that were always (or almost always) dormant, such as certain nerve cells. He extracted the nucleus, and implanted it immediately into the egg cell. Then, he put the new egg cell inside a solution which prevents it from forming a polar body (and throwing away half the genes), so that it would think it was now fertilized. It then began to divide. They had a level of success that looks like, once the technique is perfected, will approach normal external fertilization techniques. Signs of premature aging didn't occur until 5 generations of clones - this due to the fact that genes slowly change over time, and usually for the worse; we basicly extended the mouse's lifespan beyond what its DNA was designed to handle (this could be fixed by making a "DNA backup" from the original mouse, and then reconstructing that DNA each time you want a clone - an important reason for fast DNA sequencers).
The media locked onto the first story. When problems started to arise, they completely switched gears, and made it look like all cloning is dangerous. Bad media! No cookie!
-= rei =-
P.S.: BTW, I have yet to see a project where they actually transfer the mitochondria - that's over 1% of most animals' DNA.
"Well, then fire it up and show me what this..." (sigh)
They didn't per chance say this because they found 46 chromosomes in a human instead of 23 did they?
icqqm [ICQ:11952102]
The reason is that the genes don't exactly have a marker at each end to delineate them. The genes are to some extent a matter of definition. They're different lengths, they can sometimes be found in different places, and sometimes 'two genes' do exactly the same thing even though they read quite differently.
Some bacteria have been found with two or three sets of genes sort of ontop of each other- starting with different offsets. Its a bit like code that you can jump to at 0x2000 or 0x2001 and it does different, but useful things!
Anyway reverse engineering this lot will take a while...
-WolfWithoutAClause
"Gravity is only a theory, not a fact!"Uninnovate - Only the finest in engineering.
On the whole, you're right, though. The raw number of genes is more media-friendly than scientifically important. Unless your company has a business plan based on patenting genes.
Unsettling MOTD at my ISP.
It'll sort itself out over the next couple of years as the sequence gets better assembled, more non-human sequence is available for comparison and the groups adopt one another's good ideas. In the meantime, it looks like a good PR person at Ohio State managed to make their findings seem more revolutionary than they are.
By the way, if you want to bet on the number, see the GeneSweep page. (Note that bets must be placed in person!) I put my $5 on 44,000 and change.
Unsettling MOTD at my ISP.
The plain old numbers game doesn't work with the genome - we have 3 billion base pairs but frogs have 9 billion. We've got 46 chromasomes but dogs have 76. The most telling figure is that we're 99%++ genetically identical to chimpanzees and yet we can't interbreed (not that I've tried...). What really matters on a genomic level is the interplay between genes during crucial times of development, not just the functions of individual genes (our ribosomes are nearly identical to bacterial ones) - many of which were selected by evolutionary pressures meaning that once a problem had been solved, for example how to copy DNA, that gene was 'set in stone'. After all, who wants to keep solving a problem once you've found a solution? Back to the point on crucial times of development - could you imagine what the result would be if the gene(s) controlling synapse differentiation in the fetus stayed active for an extra hour? day? week? - that, ladies and gentlemen, may be all that is separating us from chimpanzees. Just a simple interplay of genes, subtly disturbed by a simple mutation, perhaps lengthening a crucial phase in brain development. So the numbers game is just irrelevant. -Nano.
It's psychology. While most scientists tend to regard the first few studies on a topic as little more than a theory until its been confirmed by a few other people, the Media and general public tend to take these as absolute answers. "Well, then" they say "Thats taken care of". If the next study indicates the first one is wrong, then somethings changed. Thats news. If it simply confirms that yes, we are little more complex than your average nematode, thats not news. If suddenly we have way more, then the number of genes in the human genome has changed(well not really but you understand what I mean) thats news and shows up in the popular media. Slashdot works the same way. Taco and company aren't going to post 14 stories confirming the number of genes - it's not new, and (to most people) not exciting. But if it challenges current beliefs, its exciting and gets posted.
As for a health skepticism about science, you should be skeptical about science. Skepticism is (or should be) an integral part of science. Nothing should be taken for granted, nothing should be accepted as true until a good number of people have had a chance to kick it around in every way they can think of without finding a problem with the theory/study/whatever. And when evidence does surface proving that the last theory was wrong, a new one should be created to fit the data and then that should be put under the microscope for flaws. This isn't religion. You don't take things on faith. Everything should be questioned and tested before acceptance.
Why?
The research article is available from the Genome Biology web site here.
Because of the way that Celera (and, AFAIK the HGP) identify genes, I feel that we are going to be seeing this soft of announcement increasingly often. Celera has been using imprinted cells as a base for their sequencing. Imprinted cell have many genes hidden or 'turned off' by interceptor protiens. What needs to be done, but could take many years, and more bio-chemistry than we have now, is to 'walk' the genome. Take the base pairs 3n at a time, and sequence, ala Folding@Home. Actual protien sequences overlap, in a sort of bio-chemical data compression. If you offset the "start here" tag by one base pair you don't always end up with a garbage protien. This is why Celera's claim of such large quantitys of 'junk' DNA was scoffed at by many in the scienfific community.
But I'm rambling. Time to go home.
The result is a lot of data, but those data are scattered in numerous databases that are organised and maintained in diverse ways by various research teams. Maybe if they open sourced their efforts and cooperated, (a) they'd spend less time potentially looking at the same things and inventing the same processes (b) somebody would get a better sense of how much has been done and how much is left to go by looking at all the data. I know, I know...there wouldn't be any patents and piles of money if they did it that way, hence they have no motivation to... And of course, then they'd then spend their time fighting off German lawsuits for naming their sequencing software KGene or some such... Gene Anderson
When Venter and Collins announced some months ago their "success" at sequencing the human genome and that according to their conclusions humans did not have much more genes that worms this did seem kind of odd to me. If this new paper published by "Genome Biology" is right it could ridicule all past celebration... i mean... if it turns out that Venter only mapped 30 000 genes but there were actually 60 000 genes in humans, wouldnt that mean that were not even half way to a complete gene map for humans??
Kind of like the comments on slashdot!
I suppose you could make it into a Katz joke too.
-PYves
So I guess we really are quite a bit more complex than worms.
Let's just forgo all the Lawyer/Management/NT Admin jokes, shall we?
Kudos, Scientists, for doubling your workload.
(I personally would say more but I didn't take Genetics due to the 8am lab.)
Personally I don't know any human Genes, although my friend had a hamster named Gene. It died, unfortunately. You'd think it would be harder to count the non-human Genes, but I guess that's far in the future. Didn't they count the mouse Genes a while back though?
"If he thinks he can hide and run from the United States and our allies, he's sorely mistaken." Bush on bin Laden
I agree with everything you've said, but I'd still like to put a bit more muscle behind the 'DNA as code' analogy. Ignoring the role that the ambient cellular environment plays would be foolhardy, true. But so would ignoring the role that hardware plays in the interpretation of software. The thing that makes the DNA the interesting part is that cells are the same from organism to organism, whereas DNA varies. Similarly with hardware and software.
It feels to me like your criticism could be leveled at a desire to understand a program by looking at the source code. "the actual branching, jumping and iteration happens in the ALU, not in the code, so the code is the wrong place to look," and so on.
Genes, I'll grant you, are the exciting bits of a chromosome, because they (generally) correspond to proteins that can be identified and detected. But, I'm not entirely clear on why they are the primary focus of genetic research.
We all seem pretty comfortable discussing DNA as though it were computer code, so let's follow that metaphor a little further. If I point at a big mess of C code (say, a console app) and ask 'What does this code do?' an amateur might be tempted to scan it for printfs, puts's, and other 'output signifiers.' But really, if that's all you look at, you don't have a clue what the actual funtion of the code is. All those boring scanf's, if/thens and operators are really important.
My rudimentary education in genetics has me convinced that DNA in a living cell has the ability (like C code) to switch, jump, branch, and (most importantly) operate recursively on its own resultant proteins. And yet, the DNA spans between genes are generally referred to as 'useless' or, in this case, 'meaningless drivel.' Am I missing something, or is this exactly where the good stuff is?
And, viewed from this angle, isn't counting genes as pointless as counting KLOCs?