When the first human genome was sequenced many people thought that all the hard work had been done. Most people thought that the genetic code, perhaps like computer code, was well organized and well laid out. But organization arises in complex systems that are designed by intelligent beings, namely ourselves. The genetic code, shaped by the slow process of evolution, is not organized. It is like a tangled mess of spaghetti code. When scientist announced that there were only 25 000 genes in the human genome, they had identified only the most organized parts of the genome, comprising a mere 1% of our DNA. Determining what the remaining 99% of the genome does will be challenging because there appear to be few organizing principles to the DNA's layout to make research easier.
This is great work. read the NYT article on the train home and then picked up a couple of the papers. The cool thing for me was the realization that when DNA is coiled, parts of the strand that are linearly very far apart end up adjacent in the coil. And further, that adjacency affects how the gene expresses proteins. That means that you not only need a decoded genome, you need to understand the relationship of the segments in 3D. For me at least it feels a lot more reasonable than the original numbers of genes. At least in terms of describing humans, as diverse and complex as we are.
Someone of course will do the same for chimps and bonobos, and then we'll know which of the switches are important as well as the genes. Very cool result and I am confident that several Nobel prizes will come out of it.
A lot more information is needed to understand the genome than just the linear sequence or 3D structure. There is phosphorylation and histone tags and RNA splicing and more (and that's just what has been discovered). All this information is needed to understand how the genome works. It is basic genetics.
According to the NYT [1] it's 'new' basic genetics. Basically all the 'junk' DNA isn't junk, and the reason the function of that DNA was not appreciated before the ENCODE effort is that these bits appeared 'far away' from the genes which it turns out they affect.
[1] "There is another sort of hairball as well: the complex three-dimensional structure of DNA. Human DNA is such a long strand — about 10 feet of DNA stuffed into a microscopic nucleus of a cell — that it fits only because it is tightly wound and coiled around itself. When they looked at the three-dimensional structure — the hairball — Encode researchers discovered that small segments of dark-matter DNA are often quite close to genes they control. In the past, when they analyzed only the uncoiled length of DNA, those controlling regions appeared to be far from the genes they affect." http://www.nytimes.com/2012/09/06/science/far-from-junk-dna-...
Obviously the 3D structure of chromosomes are important. But it is only one of many aspects that regulate our genes. While it is true we cannot understand the genome by ignoring its 3D shape, we also cannot understand the genome without looking at the many other ways the genome is regulated. Nature did not make things simple for us. There are quite a few variables that regulate our genes, and sadly it seems that all or most of these variables must be taken into account
If you knew how they found the elements together, you would be far less sanguine about the significance. They started with random sampling of close proximity of genomic positions, plus some basic structural info on how chromatins (coiled DNA molecular) are organized. Then it is a nonlinear optimization with 1e-2 convergence criteria.
Several things lay people are never told:
1. Even for fixed data, the results are variable, i.e., you will get a different answer every time you run the nonlinear programming code.
2. There are different ways of getting to 3D positions of genomes and none is ever verified.
3. Tuning parameters greatly impact results. Yet one set of tuning parameters are just as valid as another.
I had the misfortune of helping someone to extend the code for this stuff that made into a Science paper, and I can tell you right now it is the worst C++ program I have ever seen.
Is the quest for immortality the only thing that drives research like this? The PI's interviewed in the promo video for ENCODE ( http://www.youtube.com/watch?v=PsV_sEDSE2o ) can only point to biomedical implications to justify their continued funding.
This is the reality of biomedical research today. With NIH funding level flatlining in recent years and congress leery of more spending in anything other than military, medical utility is the mantra now within NIH. The irony is that most of ENCODE people are not medical researchers and are less qualified and less experienced than human genetics researchers. This is all about money. Scientists need to eat too. And reputation, and power. It is a very high-stakes game.
Agreed. While I think what ENCODE is doing is important, I feel that they oversold the value of their work. In my opinion, they obviously wanted to ensure continued funding.
It is not a bad thing that every once in a recession less fruitful areas of research are trimmed from funding to make room for more important work. To often scientist lose sight of why their research is important, and forget about trying to solve real problems. That said, too much money is poured into military research and not enough into basic science, which is what pays off in the long run.
This is the basic science that pays off in the long run. You can't just turn biomedical science projects on and off. It takes time and investment to develop the techniques and technologies to do this work, to gather the samples, and ensure data quality across the project labs. During the time that the ENCODE project was funded, the technology for doing the types of experiments to get this kind of data advanced many times. We are now talking realistically about personal genomics and the $1000 genome; at the project start we were still celebrating the 3-billion-dollar genome sequence.
Most of the people interviewed in that video are program officers/administrators at the NHGRI (an institute of the NIH) who oversaw the ENCODE project. They weren't researchers who did the work. The mission of the NIH is human health, so anything funded by them has to have a biomedical implication somewhere down the line.
I guarantee that the real impetus for the project in the researchers' eyes is to understand how an identical* instruction manual (ie., genome sequence) in every cell can give rise to a plethora of cells that do very different things to make a functioning dynamic human. In other words, for the most part, it appears that there's the sequence information and then there's how you use it. ENCODE (and modENCODE) are about how you use it.
fyi: I'm a researcher in the modENCODE consortium, a sister project to ENCODE, aimed at characterizing functional regions of DNA in two model organisms, a small worm and the fly. My PhD advisor was funded by ENCODE as well. I obviously find this stuff fascinating and important.
It used to be thought that ~95% of the genome was "junk DNA", i.e. DNA that was there either for evolutionarily historical reasons, or as the remnants of viruses or other mobile bits of DNA, but which no longer performed a significant function.
The principal finding reported by ENCODE is that the proportion of the genome which has no ascribed role has now dropped to 20%, 5% codes for "traditional" genes and 75% is this new stuff. A lot of that new stuff appears to be regulating the activity of the traditional 5%.
As to the motivations of the scientists involved, they will be a mixture of pure intellectual curiosity and the quest for cures to diseases. Of course they also would like to be paid, and must therefore dress some of their motivations in language likely to motivate their ultimate paymasters, i.e. the public.
I am an indirect member of modENCODE project -- I am paid by the project but work on computational and data analysis and therefore not wed to it. I feel I should clear up a few things.
First, that 80% figure is a guesstimate, far from definitive.
Second, those "new" discoveries are "associated" by statistical tests. All popular reporting has sidestepped the issue by not explaining the quote around word associated. Basically it means they occur together above a certain statistical threshold. For vast majority of noncoding regions we know nothing about their functions, and this situation will remain for the foreseeable future. I don't have to tell people here that correlation is not causation. Recall all the reported health benefits of wine drinking; it is same degree of certainty.
Third, new technologies inevitably have teething problems, which have not been fully worked out. This means all reported conclusions comply with accepted notions within respective scientific communities. The data are not accurate and reliable enough to overturn any received wisdom.
Forth, yes, this means the main reported scientific results are not new nor that surprising to people in the field. But this is the way popular science and specialists interact.
By and large these papers represents a lot of hard work, but in terms of amount of data and research topics they are already being overtaken, as they should be.
14 comments
[ 5.8 ms ] story [ 37.8 ms ] threadOh my god! We've sequenced and decoded the entire human genome and… we're written in _Perl!_
(And, given my gag and the backend of this site, of course I have to post: http://xkcd.com/224/ )
Someone of course will do the same for chimps and bonobos, and then we'll know which of the switches are important as well as the genes. Very cool result and I am confident that several Nobel prizes will come out of it.
According to the NYT [1] it's 'new' basic genetics. Basically all the 'junk' DNA isn't junk, and the reason the function of that DNA was not appreciated before the ENCODE effort is that these bits appeared 'far away' from the genes which it turns out they affect.
[1] "There is another sort of hairball as well: the complex three-dimensional structure of DNA. Human DNA is such a long strand — about 10 feet of DNA stuffed into a microscopic nucleus of a cell — that it fits only because it is tightly wound and coiled around itself. When they looked at the three-dimensional structure — the hairball — Encode researchers discovered that small segments of dark-matter DNA are often quite close to genes they control. In the past, when they analyzed only the uncoiled length of DNA, those controlling regions appeared to be far from the genes they affect." http://www.nytimes.com/2012/09/06/science/far-from-junk-dna-...
Several things lay people are never told: 1. Even for fixed data, the results are variable, i.e., you will get a different answer every time you run the nonlinear programming code. 2. There are different ways of getting to 3D positions of genomes and none is ever verified. 3. Tuning parameters greatly impact results. Yet one set of tuning parameters are just as valid as another.
I had the misfortune of helping someone to extend the code for this stuff that made into a Science paper, and I can tell you right now it is the worst C++ program I have ever seen.
It is not a bad thing that every once in a recession less fruitful areas of research are trimmed from funding to make room for more important work. To often scientist lose sight of why their research is important, and forget about trying to solve real problems. That said, too much money is poured into military research and not enough into basic science, which is what pays off in the long run.
I guarantee that the real impetus for the project in the researchers' eyes is to understand how an identical* instruction manual (ie., genome sequence) in every cell can give rise to a plethora of cells that do very different things to make a functioning dynamic human. In other words, for the most part, it appears that there's the sequence information and then there's how you use it. ENCODE (and modENCODE) are about how you use it.
fyi: I'm a researcher in the modENCODE consortium, a sister project to ENCODE, aimed at characterizing functional regions of DNA in two model organisms, a small worm and the fly. My PhD advisor was funded by ENCODE as well. I obviously find this stuff fascinating and important.
The principal finding reported by ENCODE is that the proportion of the genome which has no ascribed role has now dropped to 20%, 5% codes for "traditional" genes and 75% is this new stuff. A lot of that new stuff appears to be regulating the activity of the traditional 5%.
As to the motivations of the scientists involved, they will be a mixture of pure intellectual curiosity and the quest for cures to diseases. Of course they also would like to be paid, and must therefore dress some of their motivations in language likely to motivate their ultimate paymasters, i.e. the public.
First, that 80% figure is a guesstimate, far from definitive.
Second, those "new" discoveries are "associated" by statistical tests. All popular reporting has sidestepped the issue by not explaining the quote around word associated. Basically it means they occur together above a certain statistical threshold. For vast majority of noncoding regions we know nothing about their functions, and this situation will remain for the foreseeable future. I don't have to tell people here that correlation is not causation. Recall all the reported health benefits of wine drinking; it is same degree of certainty.
Third, new technologies inevitably have teething problems, which have not been fully worked out. This means all reported conclusions comply with accepted notions within respective scientific communities. The data are not accurate and reliable enough to overturn any received wisdom.
Forth, yes, this means the main reported scientific results are not new nor that surprising to people in the field. But this is the way popular science and specialists interact.
By and large these papers represents a lot of hard work, but in terms of amount of data and research topics they are already being overtaken, as they should be.