I think I could do better than 200 meg. It doesn't sound like a significant enough jump and nor is it a huge improvement over the previous proof of concept.
Plenty of companies (particularly asian ones) to long/large scale custom synthesis. The reading of the code isn't difficult these days. Assembly of the sequences into meaningful files is probably the trick.
I think a true advance would be making synthetic or highly modified polymerases which can incorporate synthetic pyrimidines/uridines into a "hyper DNA" which could improve on the 4 base nucleotide array available. Imagine a 5, 10, 26 base code. If the "DNA" doesn't have to be duplicated by an organism, it can be anything.
Imagine the packing density then.
If you put your data into an organism's "junk DNA", then the data will last forever... (barring mutations)
Very little "Junk DNA" is junk. miRNAs, gene regulatory regions, histone folding and regulation systems, chromatin folding hotspots.
Anyway, if there is true Junk DNA, it will be under less selective pressure and very quickly become riddled with mutations and errors. Moreso than protein encoding regions.
The best way to do it would to be to place an index somewhere in the genome containing references to locations within the existing coded/exome areas -- which already contain a pre-existing massive dictionary of 4-5-n base sequences.
Mapping the data would be a bitch, but if you're that desperate to encode data long term, it might be worth it. Providing your mouse/rabbit/fly/c.elegans colony doesn't die in flood/heat/poisoning/infection etc.
You're being pedantic. I've been releasing.MODs since ~1993, and then rendering them as MP3s since 1996.. (haven't been so since about 2001 now). I know of course, the format means nothing, or the distribution of sound files mean nothing.. Its the copyright violation on the work that is the problem.
I don't know how this sort of ruling will last. Where does this leave search engines? Google.com.au for example? They are just as good at finding infringing material without directly providing links, however, they might have a database where this info is stored, so how legally different is this sort of data which can provide such links, and how do judges advocate that they're any different? Other than the simple fact that the site under question clearly has a purpose for the distribution of mp3's, one wonders how far it could go?
While I totally understand your point, you're negating the simple changes that occur through mutation to provide that initial 'fitness'. The problem with GAs that you guys use, is that they too simply conceptualise what is a highly complex process. Many genetic pathways influenced by mutation take into account not only a single locus, but both loci where mutations typically have little or no effect on the net outcome of a particular protein event. Many proteins, we are finding, have a 'threshold' level of activity that is neither an 'on' or 'off' phenomena heavily utilised in modern GAs. 60% activity may be enough, and may be the result of a mutation, but a different kind of mutation could drop the activity to 45% and breach a threshold where activity is not necessary turned off, but modified. This buffering effect has a broad impact on many genetic pathways and allows the accumulation of mutations (within reason) without totally knocking the activities of these pathways out.
So while favourable recombination events could provide a fitness advantage, they are definitely not the be-all and end-all for fitness advantage. I think the estimation of mutation is grossly low, because the mutations that we find in humans usually are fortuitously identified through patients walking into clinics reporting a problem with their health, with an obvious disability or phenotype. We study those diseases, find mutated genes, and find out what has gone wrong. Contrary to this, we don't study those people of exceptional abilities, because they have no real 'problem' as such, so we don't necessarily look for genetic association between families of exceptional intelligence, or height, or of other particular traits which are generally immeasurable, e.g. a 'resistance' to lung cancer - we know some people have smoked their whole lives and have had no problems, yet others quickly succumb to the disease. There are over 35,000 genes, and well over 150,000 proteins and protein isoforms.. each has some role to play in human development and homeostasis, some are more influential than others, yet each might at some stage have the ability to allow an adaptation to a different environment. Either a 'normal' environmental adaptation, or an exceptional adaptation to something man-made.
Anyway, this is getting to the point of philosophy, so I'll leave it at that. My point is being that while GAs might give explainable results, and valid macro molecular explanations which may hold true in the earth's species, they probably underestimate the 'changes under the hood' as we're still coming to terms with the flexibility of the genome to respond to mutation, and the effect that has on development of species.
I'm a geneticist, and we typically use the terminology 'mutation' to describe any change in DNA which escapes the cell's natural ability to inhibit external of internal modification of the genome, typically using 'proofreading' mechanisms, or all the way through to cellular apoptosis to delete entire cells with certain (what the cell thinks are) detrimental mutations. Such simple changes, if they escape correction or deletion, most definitely have the ability to cause major problems, but on the flip side those mutations have the ability to modify the activities of proteins which can be advantageous (and others may do nothing at all). Whilst examples of advantageous mutations are few and far between on timescales that we can appreciate in concordance with a human lifetime, in the evolution of a species such mutations are seen to occur rather rapidly. Crossover is never a term I have come across in the definition of these terms, unless you're talking about recombination, which is another thing entirely.
Slashdot Mirror
I think I could do better than 200 meg. It doesn't sound like a significant enough jump and nor is it a huge improvement over the previous proof of concept. Plenty of companies (particularly asian ones) to long/large scale custom synthesis. The reading of the code isn't difficult these days. Assembly of the sequences into meaningful files is probably the trick. I think a true advance would be making synthetic or highly modified polymerases which can incorporate synthetic pyrimidines/uridines into a "hyper DNA" which could improve on the 4 base nucleotide array available. Imagine a 5, 10, 26 base code. If the "DNA" doesn't have to be duplicated by an organism, it can be anything. Imagine the packing density then.
If you put your data into an organism's "junk DNA", then the data will last forever... (barring mutations)
Very little "Junk DNA" is junk. miRNAs, gene regulatory regions, histone folding and regulation systems, chromatin folding hotspots. Anyway, if there is true Junk DNA, it will be under less selective pressure and very quickly become riddled with mutations and errors. Moreso than protein encoding regions. The best way to do it would to be to place an index somewhere in the genome containing references to locations within the existing coded/exome areas -- which already contain a pre-existing massive dictionary of 4-5-n base sequences. Mapping the data would be a bitch, but if you're that desperate to encode data long term, it might be worth it. Providing your mouse/rabbit/fly/c.elegans colony doesn't die in flood/heat/poisoning/infection etc.
You're being pedantic. I've been releasing .MODs since ~1993, and then rendering them as MP3s since 1996.. (haven't been so since about 2001 now). I know of course, the format means nothing, or the distribution of sound files mean nothing.. Its the copyright violation on the work that is the problem.
I don't know how this sort of ruling will last. Where does this leave search engines? Google.com.au for example? They are just as good at finding infringing material without directly providing links, however, they might have a database where this info is stored, so how legally different is this sort of data which can provide such links, and how do judges advocate that they're any different? Other than the simple fact that the site under question clearly has a purpose for the distribution of mp3's, one wonders how far it could go?
While I totally understand your point, you're negating the simple changes that occur through mutation to provide that initial 'fitness'. The problem with GAs that you guys use, is that they too simply conceptualise what is a highly complex process. Many genetic pathways influenced by mutation take into account not only a single locus, but both loci where mutations typically have little or no effect on the net outcome of a particular protein event. Many proteins, we are finding, have a 'threshold' level of activity that is neither an 'on' or 'off' phenomena heavily utilised in modern GAs. 60% activity may be enough, and may be the result of a mutation, but a different kind of mutation could drop the activity to 45% and breach a threshold where activity is not necessary turned off, but modified. This buffering effect has a broad impact on many genetic pathways and allows the accumulation of mutations (within reason) without totally knocking the activities of these pathways out. So while favourable recombination events could provide a fitness advantage, they are definitely not the be-all and end-all for fitness advantage. I think the estimation of mutation is grossly low, because the mutations that we find in humans usually are fortuitously identified through patients walking into clinics reporting a problem with their health, with an obvious disability or phenotype. We study those diseases, find mutated genes, and find out what has gone wrong. Contrary to this, we don't study those people of exceptional abilities, because they have no real 'problem' as such, so we don't necessarily look for genetic association between families of exceptional intelligence, or height, or of other particular traits which are generally immeasurable, e.g. a 'resistance' to lung cancer - we know some people have smoked their whole lives and have had no problems, yet others quickly succumb to the disease. There are over 35,000 genes, and well over 150,000 proteins and protein isoforms.. each has some role to play in human development and homeostasis, some are more influential than others, yet each might at some stage have the ability to allow an adaptation to a different environment. Either a 'normal' environmental adaptation, or an exceptional adaptation to something man-made. Anyway, this is getting to the point of philosophy, so I'll leave it at that. My point is being that while GAs might give explainable results, and valid macro molecular explanations which may hold true in the earth's species, they probably underestimate the 'changes under the hood' as we're still coming to terms with the flexibility of the genome to respond to mutation, and the effect that has on development of species.
I'm a geneticist, and we typically use the terminology 'mutation' to describe any change in DNA which escapes the cell's natural ability to inhibit external of internal modification of the genome, typically using 'proofreading' mechanisms, or all the way through to cellular apoptosis to delete entire cells with certain (what the cell thinks are) detrimental mutations. Such simple changes, if they escape correction or deletion, most definitely have the ability to cause major problems, but on the flip side those mutations have the ability to modify the activities of proteins which can be advantageous (and others may do nothing at all). Whilst examples of advantageous mutations are few and far between on timescales that we can appreciate in concordance with a human lifetime, in the evolution of a species such mutations are seen to occur rather rapidly. Crossover is never a term I have come across in the definition of these terms, unless you're talking about recombination, which is another thing entirely.