Percy's example was about increasing numbers of allelic variants leading to an increase in information in the genome. The allelic variant sequences were essentially arbitrary, they don't represent real sequences in any extant organism. This is why I say it is abstract, Percy is positing a hypothetical organism with a hypothetical gene with three then four allelic variants producing different eye colours. If this isn't abstract then can you tell me what organism it was Percy was talking about and whose biology he had so grossly incorrect?
The actual sequence is irrelevant, it isn't taken into consideration in Percy's calculations except in as much as it represents a distinct message. To criticise the biology of the sequence choice is to miss the point of the whole exercise.
Again, if it isn't abstract then what organism do you think Percy was failing to describe?
I can think of a situation where Percy's example need not be entirely biologically messed up. Suppose that the 'gene' Percy is discussing is not a protein coding gene but in fact a regulatory region upstream of a traditional protein coding gene. This protein coding gene encodes a rate limiting element of the melanin biosynthesis pathway. Let us suppose that the different eye colour phenotypes are caused by differential binding affinities to a transcription factor as a result of the small nucleotide polymorphisms at the binding site for the transcription factor, these polymorphisms would be Percy's allelic variants. This change in transcription levels of the melanin biosynthesis pathway element leads in turn to differential levels of melanin deposition in the iris.
Of course since the biology is actually irrelevant to Percy's example there is no need to do so, just as there is no need to complain about his choice of sequence.
Edited by Wounded King, : No reason given.
Edited by Wounded King, : Cleaned up pointless example.
As ever qualifications are supremely irrelevant. If Pandion has an actual specific criticism of what I wrote, other than that he didn't understand it, I have yet to make out what it is. If he just didn't understand it I'm happy to explain in greater detail.
Although I should think Percy saying what his intentions were should be sufficient to demonstrate the utter irrelevance of Drew and Pandion's discussion.
So what, if any, useful measures of information can we go by for measuring information in the genome? As Percy and Pandion show the Shannon information measure that Percy is using is essentially useless. The information content of any sequence we are studying can change at any moment unbeknownst to us as somewhere out in the wild a mutation occurs increasing the genetic diversity of that sequence within the population. The calculable information content will obviously also vary depending on the sampling I am using for my population, am I looking at all the variants in a particular species? across species? across phyla?
How do we extend this to larger scales than a single gene sequence? In a human population would every unique genome generated increase the information content in all genomes? And aren't almost all genomes unique to some small extent?
Can we ever hope to actually measure the information outside of a very specific population? Surely our incomplete knowledge would preclude us doing so?
Obviously the ID camp will never accept a concept of information where it is so trivially easy to show the widespread and continuous increase of information, even if we can't actually measure it except in controlled experimental settings.
What metrics can we use for studying a single sequence or genome? Shannon entropy? Kolmogorov complexity? Can we even make a useful measurement of information in a single gene/genome?
Is there a method for taking a pre and post mutation sequence and measuring their relative information contents?
It is interesting to me that the IDists Smooth Operator has been promoting here recently seem to be relying on sequence alignment comparisons as the basis for their information metrics rather than actual analysis of a specific sequence in isolation. Maybe isolated sequences simply aren't usable in this context.
Is it the case that what that 'information' measurement you are discussing represents is the average information we would be obtaining were we to sample a population. Therefore the more alleles are possible the less likely any particular allele is and therefore the more information we gain on finding any particular allele?
But isn't this also altered by the allele frequency?
Is your Log23=1.58 bits really 1.58 bits = -0.33 Log2(0.33) -0.33 Log2(0.33) -0.33 Log2(0.33).
So if we are talking about the de novo arrival of the new allele lets take a population of 100 and say there is 1 member with the new allele. This give us Shannon entropy =-0.33 Log2(0.33) -0.33 Log2(0.33) -0.33 Log2(0.33) -0.01 Log2(0.01) = 1.65 bits.
Am I using the wrong measurement or just measuring something completely different?
I agree that Chinese whispers is not necessarily an apt analogy, but that is principally because genetic sequences aren't like languages. The Chinese whispers analogy also leaves out the selection component which acts to eliminate elements of 'gibberish'.
Yes, and they do. The vast majority of which are so harmful that the fetus spontaneously aborts before completing gestation.
This is gross exaggeration, every newborn child has something like 30 de novo mutations distinguishing it from its parents. Were your claim to be true the levels of spontaneous abortion would be orders of magnitude higher than they actually are, even allowing for your highest figures.
Most mutations are considered neutral or 'nearly neutral'. Deleterious mutations almost certainly do considerably outweigh beneficial mutations but they are very far from constituting 'the vast majority' of mutations, and this is even less true for embryonic lethal mutations.
oh well now I have to disagree, to a certain extent - after all, natural selection favouring one mutation over the other due to a change in the breeding locale causing a shift in the allele density in a population...and that's not evolution?
He didn't say it wasn't evolution he said it wasn't an example of mutation. Of course you, me and RAZD all think that the origin of the melanic moth was a genetic mutation at some point in the species history, but so far science has not found the specific genetic basis for melanism in this instance, nor witnessed it arising de novo. So we are not in a position to definitively state that the allelic variants arose through genetic mutation, though I doubt any reasonable person would question the assumption. But if IDist/creationists were reasonable people we wouldn't keep having these same debates over and over again.
Now for the answer, and I may be wrong (I am not a geneticist but a layman), but genetics IS a language, and it already exists.
I'd tend to disagree with this. I can certainly accept that DNA can be considered a code, but I am not sure it is helpful to think of it as a language, it is certainly very far removed from being like a human language. It may fulfill several abstract properties for the philosophical conception of languages though.
you could see that as being a failed mutation - one too large to result in a working organism. Don't forget, that something like 3 out of 4 pregnancies terminates before the woman even knows she was pregnant (statstics pulled out of hat - verification anyone?).
As RAZD noted above, it is a pretty big assumption that the reason for all spontaneous abortions is an embryonic lethal mutation.
failed mutations results in a terminated pregnancy, or even no pregnancy in the first place.
This drastically reinterprets the categorisation of mutations into embryonic lethals and then everything else. This sort of 'truncation selection' is perhaps the most extreme form of selection as the allelic variant never enters the breeding pool. There are vast numbers of mutations, I'd go as far as to say the majority of deleterious mutations but I'm not 100% sure, which impair an organisms fitness but do not result in its death before it has a chance to reproduce.
A successful mutation (and every one of us is a mutated copy of our parents) results in the next generation.
Again, no. All these mutations need to be is 'not lethal'. To consider any non-lethal mutation a success, is highly idiosyncratic.
but that if these mutations are carried on then this is because they have meaning in the language.
This is where the concept of selection enters the analogy. It is selection that allows the maintenance of beneficial mutations, 'meaningful sentences' in our analogy. If our children refuse to pass on the message if it makes no sense then we will never end up with total gibberish. Of course if we just have one line of children that is game over, but consider the first child tells 2 people his message and they convey it to 2 more each who convey it to 2 more each. In this case we are much more likely to end up with a set of messages at any given point down our tree of children which all make some sense as nonsense sentences have been removed by selection.
As with the example discussed above, this is the most extreme form of selection, but it is the simplest to fit into this extended Chinese whispers analogy.
The point is (as far as i am aware) that the way the research was carried out was somewhat dodgy.
There was a thread on this some time ago, Wells' Icons of Evolution - Peppered Moths, where the evidence on this was discussed. In the scientific literature the weight of evidence still seems firmly on Kettlewell's side up to a point (Rudge, 2006). Certainly there are justifiable criticisms that a number of his field experiments were highly artificial, but there is considerable other experimental evidence to support his hypothesis that bird predation is an important selective factor on the peppered moth (Cook,2003).
As I remember It wasn't creationists debating whether or not they were stuck on trees but rather it was a debate going on betweeen everyone else. So yes, it was a bit controversial even between evolutionists.
This isn't really the case. The main proponents of the strongest 'fraud' interpretation of Kettlewell's work were Judith Hooper, a journalist, and Jonathan Wells, a vigorous creationist/ID proponent. One is a journalist rather than a scientist and the other a 'scientist' with a very determined ideological bias.
The actual scientific debate between people like Jerry Coyne and Michael Majerus, such as it was, was blown up out of all proportion by creationists and Judith Hooper's controversial book. See Majerus's own account of the matter at http://www.gen.cam.ac.uk/Research/majerus.htm by looking at either his powerpoint to the humanist association or to the ESAB in Sweden.
In terms of actual research Majerus is essentially both the harshest legitimate critic and the strongest defender of Kettlewell's work.
A point mutation changes a letter, and the whole sentance becomes meaningless, nevermind just the mutated word. How do we interpret the result of this?
This isn't really true though. A single typo may in a very strict sense make the sentence nonsense, but it is unlikely to actually stop anyone from understanding it. You may remember a well publicised bit of research a few years ago had a press release as follows ...
Aoccdrnig to rscheearch at Cmabrigde uinervtisy, it deosn’t mttaer waht oredr the ltteers in a wrod are, the olny iprmoetnt tihng is taht the frist and lsat ltteres are at the rghit pclae. The rset can be a tatol mses and you can sitll raed it wouthit a porbelm. Tihs is bcuseae we do not raed ervey lteter by it slef but the wrod as a wlohe.
Clearly this is technically nonsense but I very much doubt that you can't understand what it says.
If the mutation is in an instruction code binary string, it will simply change it to another instruction and something will still happen. With DNA point mutations in coding sections, amino acids will still be selected, proteins will still be built, things will still happen - whatever they happen to be...
Would you agree?
I definitely agree with this description. To try and abstract the actual function/information from the biochemistry of the actual molecules is to lose sight of what is actually happening in reality. This is what leads to views like Drews and Smooth operators based on the assumption that ...
The information inside DNA must be specified with little margin for error.
When for the most part we have no ideas what the allowable margins for error are because we don't know what the functional effects of most possible mutations on a protein/sequence would be.