I'm sure it isn't (there was an article about this in the Economist, www.economist.co.uk, a few weeks back) approaching a majority in a linear sense; what they mean is that if it keeps doubling at the same rate (like, 4 months, as I recall) it'll be a majority pretty quickly. So, it's approacing a majority at an exponential rate. You could view this as "quickly approaching" a majority if you squint hard enough.
Sam
Re:Wasn't this technology invented in the 70's?
on
Bioinformatics
·
· Score: 1
Much of it was, yes.
The sequence recognition software (i.e. BLAST; someone should post a link) is based on 1970s voice recognition technology - somebody's table algorithm. Neural Nets, which are fairly popular for a diversity of applications other than sequence recognition, were also invented in the 70's, if I recall.
Markov Chain Monte Carlo (a means of solving high dimensional integration through random sampling) was invented by some of the physicists who worked on the hydrogen bomb in the 1950s, but it was only recently (circa 1988) adopted by Statisticians, in particular in Biostat.
Much of this is fairly old technology but it needs significant retooling to be applied to the problems of biology, which are very, er.... rich.
In any case, most bioinformatics is, in fact, not ground breaking in computational terms. There is a huge amount of work to be done figuring out how to take advantage of existing information technology, ESPECIALLY in the area of Database design.
Bioninformatics is one of the best fields to get involved in as an academic. There is significant private sector interest - which means that jobs are available and that there is always teaching demand to keep you employed - but there isn't as much bullshit flying as in really corporately penetrated fields, like nutrition (which I used to do.)
Sam,
presently UCSC biochemistry + applied math, start to get my belated PhD at Columbia in the Fall. Yippee! Praetorious is a character from Frankenstein.
Fools! The real point of this tecnhology is to bury nuclear devices deep beneath the earth's crust; the present nuclear payload of the US (which I'm sure all slashdotters agree we want to largely dispose of anyway, yes?) is _more_ than enough to dig a hole, half a mile across, all the way through the mantle, IF you can get the bombs anywhere you want. Which, thanks to this technology, you can.
Now, the process will release huge amounts of vaporised rock, I grant. Oxidised rock, by the way, SiO2, is a hugely potent greenhouse gas - at temperatures around the surface of the sun! SiO2 is _quartz_ people. It is NOT A GAS. The borehole will also release large quantities of greenhouse gases, but mostly it will release dust and ash; countering the greenhouse effect.
Anyway, you dig it in baha california, below sea level, and then you run a trench to it. You build turbines over the hole and collect the water (for use in agriculture.) The electricity from the borehole, about half of it, goes to maintaining a nearby smelting plant which has to recast the turbines continuously (as they're coated in heavy metal deposits) but you still make an electric profit. A HUGE one. AND you make a great deal of agricultural grade water (no, I'm not expecting people to drink it.)
I don't think that anything which is designed can qualify as life - genetically modified things are, by the way, very far from being designed (I'm a geneticist.) The really fascinating property of real living things, as they exist in nature, is that they are not designed or streamlined - they are endlessly refreshing and unexpected, they have properties which you could not concieve of, acquired in response to environmental factors of which you are not even aware. Even an artificial creation which you place in a highly complex environment and permit to "evolve" is really just echoing your own ideas and preconceptions back at you - an environment designed by more than one person is more along the lines of what I'm thinking about but is still qualitatively different from the real world in a way that precludes, to my thinking, containing anything which is alive.
Consciousness, however, you can have. We think it's such a cool trick just because we do it but the mixed scum living under twenty cubic meters of top soil is, as an aggragate, far more complex than a human being, possessed of more of this quality that I think seperates actual life from this digital stuff, and nothing like sentient.
Sam
I just got into the columbia university biology graduate program. I will do my victory dance.
Because it prevents the movement in opposition to people owning ideas from being owned itself. Napster is a for-profit corporation, and (we all recall the deal with Bartelsman) has zero ideological commitment to anything. It's loss does NOT (as another poster asserted) mean the death of p2p; it means the death of p2p as a commercially viable enterprise - good.
This means that peer-to-peer is going to move toward the control of the peers involved, which is good, right and proper. Now that napster is out of the way (and benefiting from the publicity that napster has generated) we'll start to see a real explosion of creative alternatives, some of which are going to be impossible to police, legal or not.
Yeah, I'm being premature. Sue me:).
Billions and billions and billions of theories - some of which are going to be right.
Look, I love Carl Sagan, but I think it's a little overboard to discount the opinion of anyone who hasn't read his book(s). The point could be made that this slashdot thing exists purely for people to express their opinions, however misinformed they may be, because the opinions of people are important in their own right. I'd suggest that you won't much enjoy discussion groups if that isn't your attitude.
Finally, if mere bioinformaticians such as myself are incapable of understanding the complexities of the arguments why it is "impossible" for formations of this sort to arise by abiotic means, I'd be very suspicious of said arguments. Our understanding of physical chemistry under the conditions that existed inside that meteorite is so poor that I think it's very, very premature for you to conclude that no abiotic explanations exist for the formation of those structures. That said, I agree that the evidence is fairly persuasive but it's obvious to me that a LOT more research needs to be done before we can say anything conclusive.
Don't swear at street lunatics. 's rude.
It is _possible_ that other sorts of galaxies (which might, for poorly understood reasons, have experienced greatly more, greatly less, or a different distribution of supernovae) would have greatly higher levels of ambient heavy metals or radio-isotopes, which might very well prevent life from arising, or insufficient heavy metals to have formed earth-size planets. Hardly seems a given, but it's possible.
No "more temperature tolerant life" might be possible. Differential levels of radiation throughought a sun's life-cycle might make other luminosity-distance combinations impossible; also, other solar effects might prevent formation of planets at appropriate distances to blue giants and suchnot.
The reactions that cause life to form might never occur under 10 Gs; it's entirely possible that they absolutely require liquid water; pressure at sea level could be quite massive (although that would be expected to favor catabolic rxns, can't say) possible also liquid water with certain other features - temperature, concentrations of ions and sapponins (er.. soaps) and so on.
The moon has been suggested as a requisite for the formation of _complex_ life because of it's role as an asteroid-sweeper.
What we do know is that planets we have yet detected do not generally have circular orbits like those in our solar system. That doesn't bode well for high frequency extra-terrestrial life, anyway you slice it. Neither does the detection of such a large number of Jupiter-class objects within 1 AU of their respective primaries.
That said, even if first-poster is right, they are billions and billions and billions (Carl Sagan admits to saying millions but says he never said that, btw) of stars; you do the math.
You can patent the sun, the moon and the stars but it isn't going to do you a lot of good.
Patenting an actual gene (unless, like the ones monsanto wants to patent, it has been modified) is possible but not ENFORCABLE. You can force anyone else who wants to use the gene to go through an expensive legal hurdle (which is Monsanto's actual tactic with the whole gene patenting thing) but you can't actually stop them unless the gene represents an invention in even the most loosely defined sense.
For example, my family has a patent on using artificial eggs to deliver fat soluble nutrients (this isn't a gene but it's the same law.) Does this apply to actual, natural eggs? No. Is it an enforcable patent? Yes. Some guy in australia tried to patent any mixture of anything used to deliver carotenoids; this patent, while he does have it, is not enforcable.
So, yeah, they could patent your genes, and get squat. They could patent a laboratory-modified form of your gene, and that would probably stick. They could patent the *innovation* of *cloning* your gene into some other organism, and that, absurd as it is, will stick. They don't own your gene, they just own the innovation of using it transgenically; and their patent will *probably* only stick, however they worded the application, to someone else trying to clone the gene for the same, and I know judges don't know genetics from deer shit, "purpose."
All of this is totally seperate from the question of biopharmaceutical patents - the *vast* majority of genetic type patents are, and will continue to be, agricultural. You're not going to see heavy medical genetic patenting until some fool convinces some other fools to start injecting people with "therepeutic" retroviruses; which is a VERY VERY STUPID (TM) thing to do.
Actually, there was a sequel to Master of Magic. It was called "Age of Wonders," it wasn't done entirely by the same people and it didn't have klackons in it, but it was clearly intended to be a sequel. Quite frankly, it also sucked - it was warlords everywhere MOM was civilisation.
That said, they should certainly make a sequel to MOM. I'm sure if MOO III sells well, they will.
Yeah, the human genome is several gigs; but the vast majority of it isn't "coding." If you were going to present the whole human genome (as opposed to, more realistically, a short sequence of particular interest to your research) you'd be able to convey a LOT more information by presenting the 1% of it that codes for amino acids, along with markups to provide links to crystal structures of the proteins, little sub-charts showing the frequency of medically relevant site specific substitutions (recall, there isn't _a_ human genome, there are many different ones) and so on and so forth. Yeah, it might blow up past the size of the raw genome, but it would contain actually useful information.
That said I can't think of any features you could want in such a language you can't do with just old html. Shrug.
Firstly, it's _not_all that much information. 1 mega-base = 2 mega-bytes; DNA only comes in A,C,T and G.
The search enginge that makes it possible for scientists around the world to coorelate this sort of data is called BLAST:
http://www.ncbi.nlm.nih.gov/BLAST/
If you poke around the web-site for a while, somebody has probably come up with something better than BLAST while I wasn't paying attention.
I really should learn html. Anybody recommend a particular book (as opposed to the doznes of others, not just one that you happen to know exists.)
The human genome has a lot of things in it called "Introns." They are, in a terrible analogy, like comments in code. Like comments, they *may* have subtle effects on who uses the code - the mRNA (which is what the gene turns into on it's way to becoming a protein) spends a while with the Introns still in it, and these Introns may have some effect on the frequency or likelihood of the mRNA in question eventually being used to make proteins. Vast oversimplification, of course. "Lower" organisms also do something called "alternative splicing" with their introns, where you occasionally skip the coding region inbetween two introns, and thus the same gene ends up leading to two different mRNAs with some frequency.
The genes which are "found" in fugu are found by homology - in particular, they run the human gene and the fugu gene sequences through some sort of state machine with weighted transitions (it's the same technology that the first voice recognition stuff uses, if any of you do that) - if the total weight of the transitions in the best possible traversal (this is determined by some kind of dyanmic algorithm the name of which I cannot remember) is below a certain amount, it means (and yes, it really does mean this) that thet two genes are evolutionarily related. The name of the engine in question, which searches a database for genes that score high on this sort of test, is BLAST.
This may or may not mean that the genes code for proteins with similar functions - it probably does, but you can't say for certain.
Obviously, the Puffer fish is going to have many genes that a human being doesn't, however, the vast majority of genes are going to be, at the very least, homologous. A person "shares" 50% of her genes with a banana; i.e. the genes are present in homologous form.
Recall, human beings have gills for a significant period of time. Whatever genes encode for the signals that causes gills to form - and that is the way it works, genes encode chemicals, there is no gene for gills - are present in a person, and if we can find them in a puffer fish it becomes much, much easier to locate them in a human genome. Furthermore, once we know what the genes that make the signals that end up forming gills are - it becomes easier to find the (presumably similar) genes that signal for hair, fingernails, lungs and so on.
Finally, the Genes that are in a person and not in a puffer are probably NOT the point of real interest in the human genome. The big differnce between a human being and a puffer is not in what genes are actually present, but in 1. sequence differences between homologous genes in the puffer and in a person, and 2. in differences between activity (transcriptional frequency) of homologous genes. It is this difference in the rate at which a particular gene is transcribed (which can, in some cases, be traced to a totally different gene, but I digress) which cause you to no longer have gills like you did when you were -1/2 years old.
The fundamental stuff - the stuff that both a human and a fugu have in common - is still nothing like fully understood. The analogy to the main fxn of a c-program is invalid; the developmental genes are more like the symbol table of the c-compiler, which is the first thing you'd want to take apart if you wanted to reverse-engineer how c works from some very poorly commented code.
Slashdot Mirror
I'm sure it isn't (there was an article about this in the Economist, www.economist.co.uk, a few weeks back) approaching a majority in a linear sense; what they mean is that if it keeps doubling at the same rate (like, 4 months, as I recall) it'll be a majority pretty quickly. So, it's approacing a majority at an exponential rate. You could view this as "quickly approaching" a majority if you squint hard enough.
Sam
Much of it was, yes.
The sequence recognition software (i.e. BLAST; someone should post a link) is based on 1970s voice recognition technology - somebody's table algorithm. Neural Nets, which are fairly popular for a diversity of applications other than sequence recognition, were also invented in the 70's, if I recall.
Markov Chain Monte Carlo (a means of solving high dimensional integration through random sampling) was invented by some of the physicists who worked on the hydrogen bomb in the 1950s, but it was only recently (circa 1988) adopted by Statisticians, in particular in Biostat.
Much of this is fairly old technology but it needs significant retooling to be applied to the problems of biology, which are very, er.... rich.
In any case, most bioinformatics is, in fact, not ground breaking in computational terms. There is a huge amount of work to be done figuring out how to take advantage of existing information technology, ESPECIALLY in the area of Database design.
Bioninformatics is one of the best fields to get involved in as an academic. There is significant private sector interest - which means that jobs are available and that there is always teaching demand to keep you employed - but there isn't as much bullshit flying as in really corporately penetrated fields, like nutrition (which I used to do.)
Sam,
presently UCSC biochemistry + applied math, start to get my belated PhD at Columbia in the Fall. Yippee! Praetorious is a character from Frankenstein.
Fools! The real point of this tecnhology is to bury nuclear devices deep beneath the earth's crust; the present nuclear payload of the US (which I'm sure all slashdotters agree we want to largely dispose of anyway, yes?) is _more_ than enough to dig a hole, half a mile across, all the way through the mantle, IF you can get the bombs anywhere you want. Which, thanks to this technology, you can.
Now, the process will release huge amounts of vaporised rock, I grant. Oxidised rock, by the way, SiO2, is a hugely potent greenhouse gas - at temperatures around the surface of the sun! SiO2 is _quartz_ people. It is NOT A GAS. The borehole will also release large quantities of greenhouse gases, but mostly it will release dust and ash; countering the greenhouse effect.
Anyway, you dig it in baha california, below sea level, and then you run a trench to it. You build turbines over the hole and collect the water (for use in agriculture.) The electricity from the borehole, about half of it, goes to maintaining a nearby smelting plant which has to recast the turbines continuously (as they're coated in heavy metal deposits) but you still make an electric profit. A HUGE one. AND you make a great deal of agricultural grade water (no, I'm not expecting people to drink it.)
Sam
I don't think that anything which is designed can qualify as life - genetically modified things are, by the way, very far from being designed (I'm a geneticist.) The really fascinating property of real living things, as they exist in nature, is that they are not designed or streamlined - they are endlessly refreshing and unexpected, they have properties which you could not concieve of, acquired in response to environmental factors of which you are not even aware. Even an artificial creation which you place in a highly complex environment and permit to "evolve" is really just echoing your own ideas and preconceptions back at you - an environment designed by more than one person is more along the lines of what I'm thinking about but is still qualitatively different from the real world in a way that precludes, to my thinking, containing anything which is alive.
Consciousness, however, you can have. We think it's such a cool trick just because we do it but the mixed scum living under twenty cubic meters of top soil is, as an aggragate, far more complex than a human being, possessed of more of this quality that I think seperates actual life from this digital stuff, and nothing like sentient.
Sam
I just got into the columbia university biology graduate program. I will do my victory dance.
Because it prevents the movement in opposition to people owning ideas from being owned itself. Napster is a for-profit corporation, and (we all recall the deal with Bartelsman) has zero ideological commitment to anything. It's loss does NOT (as another poster asserted) mean the death of p2p; it means the death of p2p as a commercially viable enterprise - good. :).
This means that peer-to-peer is going to move toward the control of the peers involved, which is good, right and proper. Now that napster is out of the way (and benefiting from the publicity that napster has generated) we'll start to see a real explosion of creative alternatives, some of which are going to be impossible to police, legal or not.
Yeah, I'm being premature. Sue me
Billions and billions and billions of theories - some of which are going to be right.
Look, I love Carl Sagan, but I think it's a little overboard to discount the opinion of anyone who hasn't read his book(s). The point could be made that this slashdot thing exists purely for people to express their opinions, however misinformed they may be, because the opinions of people are important in their own right. I'd suggest that you won't much enjoy discussion groups if that isn't your attitude.
Finally, if mere bioinformaticians such as myself are incapable of understanding the complexities of the arguments why it is "impossible" for formations of this sort to arise by abiotic means, I'd be very suspicious of said arguments. Our understanding of physical chemistry under the conditions that existed inside that meteorite is so poor that I think it's very, very premature for you to conclude that no abiotic explanations exist for the formation of those structures. That said, I agree that the evidence is fairly persuasive but it's obvious to me that a LOT more research needs to be done before we can say anything conclusive.
Sam
Don't swear at street lunatics. 's rude.
It is _possible_ that other sorts of galaxies (which might, for poorly understood reasons, have experienced greatly more, greatly less, or a different distribution of supernovae) would have greatly higher levels of ambient heavy metals or radio-isotopes, which might very well prevent life from arising, or insufficient heavy metals to have formed earth-size planets. Hardly seems a given, but it's possible.
No "more temperature tolerant life" might be possible. Differential levels of radiation throughought a sun's life-cycle might make other luminosity-distance combinations impossible; also, other solar effects might prevent formation of planets at appropriate distances to blue giants and suchnot.
The reactions that cause life to form might never occur under 10 Gs; it's entirely possible that they absolutely require liquid water; pressure at sea level could be quite massive (although that would be expected to favor catabolic rxns, can't say) possible also liquid water with certain other features - temperature, concentrations of ions and sapponins (er.. soaps) and so on.
The moon has been suggested as a requisite for the formation of _complex_ life because of it's role as an asteroid-sweeper.
What we do know is that planets we have yet detected do not generally have circular orbits like those in our solar system. That doesn't bode well for high frequency extra-terrestrial life, anyway you slice it. Neither does the detection of such a large number of Jupiter-class objects within 1 AU of their respective primaries.
That said, even if first-poster is right, they are billions and billions and billions (Carl Sagan admits to saying millions but says he never said that, btw) of stars; you do the math.
You can patent the sun, the moon and the stars but it isn't going to do you a lot of good.
Patenting an actual gene (unless, like the ones monsanto wants to patent, it has been modified) is possible but not ENFORCABLE. You can force anyone else who wants to use the gene to go through an expensive legal hurdle (which is Monsanto's actual tactic with the whole gene patenting thing) but you can't actually stop them unless the gene represents an invention in even the most loosely defined sense.
For example, my family has a patent on using artificial eggs to deliver fat soluble nutrients (this isn't a gene but it's the same law.) Does this apply to actual, natural eggs? No. Is it an enforcable patent? Yes. Some guy in australia tried to patent any mixture of anything used to deliver carotenoids; this patent, while he does have it, is not enforcable.
So, yeah, they could patent your genes, and get squat. They could patent a laboratory-modified form of your gene, and that would probably stick. They could patent the *innovation* of *cloning* your gene into some other organism, and that, absurd as it is, will stick. They don't own your gene, they just own the innovation of using it transgenically; and their patent will *probably* only stick, however they worded the application, to someone else trying to clone the gene for the same, and I know judges don't know genetics from deer shit, "purpose."
All of this is totally seperate from the question of biopharmaceutical patents - the *vast* majority of genetic type patents are, and will continue to be, agricultural. You're not going to see heavy medical genetic patenting until some fool convinces some other fools to start injecting people with "therepeutic" retroviruses; which is a VERY VERY STUPID (TM) thing to do.
shandelm@cats.ucsc.edu
Actually, there was a sequel to Master of Magic. It was called "Age of Wonders," it wasn't done entirely by the same people and it didn't have klackons in it, but it was clearly intended to be a sequel. Quite frankly, it also sucked - it was warlords everywhere MOM was civilisation.
That said, they should certainly make a sequel to MOM. I'm sure if MOO III sells well, they will.
Sam
Yeah, the human genome is several gigs; but the vast majority of it isn't "coding." If you were going to present the whole human genome (as opposed to, more realistically, a short sequence of particular interest to your research) you'd be able to convey a LOT more information by presenting the 1% of it that codes for amino acids, along with markups to provide links to crystal structures of the proteins, little sub-charts showing the frequency of medically relevant site specific substitutions (recall, there isn't _a_ human genome, there are many different ones) and so on and so forth. Yeah, it might blow up past the size of the raw genome, but it would contain actually useful information.
That said I can't think of any features you could want in such a language you can't do with just old html. Shrug.
UCSC Molecular Biology
Firstly, it's _not_all that much information. 1 mega-base = 2 mega-bytes; DNA only comes in A,C,T and G.
The search enginge that makes it possible for scientists around the world to coorelate this sort of data is called BLAST:
http://www.ncbi.nlm.nih.gov/BLAST/
If you poke around the web-site for a while, somebody has probably come up with something better than BLAST while I wasn't paying attention.
I really should learn html. Anybody recommend a particular book (as opposed to the doznes of others, not just one that you happen to know exists.)
Dr. Praetorious
UCSC Molecular Biology
The human genome has a lot of things in it called "Introns." They are, in a terrible analogy, like comments in code. Like comments, they *may* have subtle effects on who uses the code - the mRNA (which is what the gene turns into on it's way to becoming a protein) spends a while with the Introns still in it, and these Introns may have some effect on the frequency or likelihood of the mRNA in question eventually being used to make proteins. Vast oversimplification, of course. "Lower" organisms also do something called "alternative splicing" with their introns, where you occasionally skip the coding region inbetween two introns, and thus the same gene ends up leading to two different mRNAs with some frequency.
The genes which are "found" in fugu are found by homology - in particular, they run the human gene and the fugu gene sequences through some sort of state machine with weighted transitions (it's the same technology that the first voice recognition stuff uses, if any of you do that) - if the total weight of the transitions in the best possible traversal (this is determined by some kind of dyanmic algorithm the name of which I cannot remember) is below a certain amount, it means (and yes, it really does mean this) that thet two genes are evolutionarily related. The name of the engine in question, which searches a database for genes that score high on this sort of test, is BLAST.
This may or may not mean that the genes code for proteins with similar functions - it probably does, but you can't say for certain.
Dr. Preatorious
UCSC Biology
Obviously, the Puffer fish is going to have many genes that a human being doesn't, however, the vast majority of genes are going to be, at the very least, homologous. A person "shares" 50% of her genes with a banana; i.e. the genes are present in homologous form.
Recall, human beings have gills for a significant period of time. Whatever genes encode for the signals that causes gills to form - and that is the way it works, genes encode chemicals, there is no gene for gills - are present in a person, and if we can find them in a puffer fish it becomes much, much easier to locate them in a human genome. Furthermore, once we know what the genes that make the signals that end up forming gills are - it becomes easier to find the (presumably similar) genes that signal for hair, fingernails, lungs and so on.
Finally, the Genes that are in a person and not in a puffer are probably NOT the point of real interest in the human genome. The big differnce between a human being and a puffer is not in what genes are actually present, but in 1. sequence differences between homologous genes in the puffer and in a person, and 2. in differences between activity (transcriptional frequency) of homologous genes. It is this difference in the rate at which a particular gene is transcribed (which can, in some cases, be traced to a totally different gene, but I digress) which cause you to no longer have gills like you did when you were -1/2 years old.
The fundamental stuff - the stuff that both a human and a fugu have in common - is still nothing like fully understood. The analogy to the main fxn of a c-program is invalid; the developmental genes are more like the symbol table of the c-compiler, which is the first thing you'd want to take apart if you wanted to reverse-engineer how c works from some very poorly commented code.
Dr.Praetorious
UCSC Molecular Biology