GENETICS- CREATIONISTS need not be INTIMIDATED.

by hooberus 40 Replies latest watchtower beliefs

  • hooberus
  • bohm
    bohm

    hooerus, BTS: Sorry i have not answered, i have been very busy. I will return when i have read the articles in question.

  • Terry
    Terry

    Here is my problem (and it is insurmountable for me) with the debate between religion and science.

    I DON'T KNOW ENOUGH to take sides!!

    There, I've said it.

    Let me back up a moment and explain.

    When I frst encountered the Jehovah's Witness theology as a 14 year old teenager my discussion consisted of two things:

    1.Assertions from my friend who backed up his assertions with Scripture reading.

    2. Common sense (such as I possessed) objections and counter arguments.

    That is all I had in my critical thinking, skeptical arsenal.

    I knew NOTHING about the bible. I had may as well try to argue with a Mathematics Professor about asymptotes for all I knew.

    There has to be a Fundamental Base of Knowledge on the listener's part to decide an argument based on information and data.

    When it goes beyond that base you are in no-man's-land.

    You can't know more than you know just as you can't lift yourself by your own bootstraps.

    Having said all the above; I say this.

    Only Professors, Scientists and experts can argue on the technical level. The rest of us are easily swayed by our Presuppositions into a Confirmation Bias.

    So, I opt out. I just don't know what I'm talking about.

  • bohm
    bohm

    The problem with Sandford

    Hello! I have read a couple of papers on Sanfords model of evolution (Mendels Accountant) where he test his hypothesis that darwins first axiom is wrong.

    From my reading of his papers this is what his program does:

    First it assign a fitness (initially 1) to each individual in the population and a genomen. The fitness indicate how likely the individual is to replicate, and how many offsprings it get.

    Here is the thing: We have already left the land where the simulation is biologically realistic.

    The world does not work like that. Ask yourself what the fitness of a 'dog' is. Is the fitness of a 'dog' larger than a cat? An E-Coli? What does it mean the fitness increase or decrease by a small amount? how does that look?

    In reality fitness is a function of the world. Mutations that are bad in one context are good in others -longer legs may be good in Africa, but bad in Greenland. Im not saying this invalidate the ability of the model to say something about the world, not at all! Almost all models of the world rely on strong assumptions, im just saying that we are throwing out a *lot* of complexity, and at some point we need to check if we threw out something essential.

    ANYWAY. The fitness of the animal are in his simulation affected by a variety of factors, the most important is mutation. Mutations are small changes to the genomen. Each change is assigned a 'change in fitness', just a small random number, which is drawn from a distribution and alters the total fitness of the individual. Here is the crucial thing and where many feel his model is wrong:

    He assume per default there is a much higher likelihood of a 'bad' (negative) change than a 'good' (positive),

    in particular the ratio is 0.0001, but whats even worse, he assume that given if it is a good or bad mutation,

    the fitness effect is found through a Wiebull function that is furthermore SCALED by a paremeter d_sf = -1 (bad mutations) and d_sf = 0.01 (good). Otherwise the two Wiebull functions use the same parameters.

    Okay in laymans terms: He first state there are 10000 more bad than good mutations. Then he state that the change in fitness of a bad mutation is on average 100 times stronger than good.

    This has the following effect: There will be a gazillion 'bad' mutations that individually affect the overall fitness very little, but will always draw downwards, and often the bad mutations will kill the individual directly (fitness 0). Very rarely there will be good mutations, but they will have so little effect to the total

    fitness that they are very easily drowned in 'noise' and hence cannot be selected for.

    It is clear to anyone that with a finite population size (typical size in his simulations is 1000) the effect will be a population crash. How quickly depends on the input, but an order of 300 generations for a human population of 1000. Sanford has on various conferences used this to support the claim that humans has been here in less than 300 generations, and claim that the reason they lived for 500 years right around the flood was because their 'fitness' was higher.

    Why this is bullshit:

    One can make many theoretical objections to Sanfords model, or indeed just show him examples that the very ting he says his program show is impossible, that positive mutations are selected for, actually happends in nature. Here are a couple of links:

    http://www.nature.com/nature/journal/v413/n6855/full/413514a0.html

    http://www.nature.com/nature/journal/v419/n6909/full/nature01140.html

    http://biology.plosjournals.org/perlserv/?request=get-document&doi=10.1371/journal.pbio.0030170

    http://www.nature.com/nature/journal/v437/n7055/full/nature04072.html

    http://biology.plosjournals.org/perlserv/?request=get-document&doi=10.1371/journal.pbio.0040072

    http://scienceblogs.com/evolgen/2006/03/detecting_natural_selection_pa.php

    And this is where Sanfords is so badly wrong. Science is about describing the nature, and when you want to evaluate an add-hoc model like Sanfords you got to evaluate it based on how well it describe nature. ALL models, no matter how good they look, have to undergo that test. I write computer simulations of statistical models that describe systems much, much simpler than what Sanford try to describe, and i have NEVER read an article that did not contain validation against ACTUAL DATA FROM THE REAL WORLD.

    So how does Sanfords model describe a population of E-coli? of the mice on the ark? Fruitflies? Sanford are very silent of those systems, and the reason is clear: E-coli would go extinct several times a year with his model of evolution, so clearly his model is not biologically realistic.

    Thats what i mean by validation. The stuff you linked to was only validation that his program does what he sayes it does (not that it matters, but after the source was made puplic it was found it was riddled with bugs, but lets not go there), not that his model actually describe the essentials of evolution. I can easily make a program that validate evolution, heck, lets write it right now:

    1) Initialize a population of 1000 individuals, each with a fitness of 1.

    2) (mutation) Add a number uniformly distributed from (-1, 1) to each individuals fitness.

    3) (replication) Delete the 500 individuals with lowest fitness. Copy them to get to 1000 again. Go to 2).

    but that does not prove jack shit because i have not validated it against reality.

  • Gerard
    Gerard

    Terry, your statement is sober, and if you had to decide your next move based on the information available, what is more reliable and real?: Scientific consensus (e.g. knowledge) (evolution, molecular genetics, geology, paleontology, physics, etc) or faithfull hopes?

    The existence of god or an afterlife is a complete unknown to both science and religion, so it is fair to talk about beliefs or faith but without claiming the truth. I am a scientist and I have a gut feeling that there is an afterlife. I will find out when i die, and I will accept it if that is the grand cycle of Qi energy, I just don't rely on someone else trusting it is that way. I have a problem against religion when they thrust creationism in schools because not one religion has been able to show one piece of evidence to support their product confident belief in a parallel reality.

    Creationism is nothing more than Judeo-Christian faith without logical proof or material evidence. It offers no facts while demanding to disregard the very concept of established knowledge.

  • bohm
    bohm

    BTS: Thank you for your long answer - it allmost deserve a topic of its own!

    First i got to admit i am in over my head on the biological side here. I dont really know a lot about biology, it is not my field and i have not even read an introductuary textbook. The way evolution has been an interest of me is from an information-theoretic perspective: The genomen clearly contain 'information' (though it is hard to quantify exactly how much), and the process of life can be seen as a chanel of communication from now and n generations forward, and what happends in between - selection, mutation, horizontal gene transfer, etc, lets call them all 'process E' - is what creationist and evolutionist is talking about: Is it possible that these processes may increase the amount of 'information', or does it just work as a noise filter that decrease it?

    So having stated my ignorance i think you will understand why i have properly misunderstood you. The way i read your original post was that all DNA in the genomen was there for a reason, ie. it was like a house of card: If we removed anything, the whole thing would tumble down and you wouldnt get a human. I think the misunderstandment come from how i understand the term 'junk-dna'.

    When you and i can taste sweet things its because a part of our DNA code for a specific protein that allow us to do so. In cats the situation is different. That gene has a point mutation so that it does no longer 'work', ie. now it does not produce any proteins and the cat cannot taste sugar.

    Thats what i mean by 'junk'. A region that serve no function in th sence you can pretty much flip all the 'bits' in the dna string and you would still get a cat. The damaged gene /may/ still have a weak function (structually), and there may be some constraints on how many bits you can flip, but those restrictions is much, much weaker than on the other parts of the genomen (a way to say that is the entropy is higher), but pretty much all random flips of this part of the genomen should have no effect on the cat.

    The reason that i expect such 'junk' is properly a bit strange for a biologist. As i have written many times over 'evolution' (or genetic algorithms) is a well-known optimization strategy in engineering, and i believe what works and what does not work when you work with genetic algorithms can tell us something about what to expect from our DNA. If you take down a textbook on how you build a genetic algorithm it will properly give you the following guidelines:

    "Your representation (genomen) must be very redundant and as large (wastefull) as possible". So thats what i expect from our genomen: That there are many duplicated, nonfunctional or partially nonfunctional structures, and that the functional structures are often coded in a very wastefull manner; that is, junk dna.

    Actually i think a program such as dr. Sanfords can say something about this situation. In a bacteria a much larger DNA (full of 'true' junk) would be bad, because building the dna is a very expensive thing compared to the other stuff the bacteria does. Because of this the 'junk' factor in with a higher relative value in the bacterias 'fitness function', and because there is a huge population of bacteria they are under a more fine-grained selection function and hence in bacterie it is possible for evolution to weed out 'junk'.

    In humans it is different. Our fitness function is much, much more advanced because we consist of 40 billions cells and interact with a much more complicated enviroment. Furthermore there are allmost none of us compared to bacteria. Therefore i would expect evolution would have a much harder time to weed out 'junk' in our genomen, and thus i would expect humans (on average) to contain more junk propertionally compared to a bacteria strain, given the same initial conditions.

  • hooberus
    hooberus
    The problem with Sandford
    Hello! I have read a couple of papers on Sanfords model of evolution (Mendels Accountant) where he test his hypothesis that darwins first axiom is wrong. . .

    Why this is bullshit:
    One can make many theoretical objections to Sanfords model, or indeed just show him examples that the very ting he says his program show is impossible, that positive mutations are selected for, actually happends in nature. Here are a couple of links:
    http://www.nature.com/nature/journal/v413/n6855/full/413514a0.html
    http://www.nature.com/nature/journal/v419/n6909/full/nature01140.html
    http://biology.plosjournals.org/perlserv/?request=get-document&doi=10.1371/journal.pbio.0030170
    http://www.nature.com/nature/journal/v437/n7055/full/nature04072.html
    http://biology.plosjournals.org/perlserv/?request=get-document&doi=10.1371/journal.pbio.0040072
    http://scienceblogs.com/evolgen/2006/03/detecting_natural_selection_pa.php

    Actually his program allows for the selection of positive mutations. As far as the links that you provided, the main problem with them is that they often start with the assumption that humans, chimps, and other apes share a common ancestor to begin with, and that humans evolved from this ape like ancestor over million of years. (the assumption of the evolution in question), Then based on this primary assumption and other secondary evolutionary assumptions attempt to detect “positive selection”. Even these secondary assumptions are questionable.

    http://www.discovery.org/a/14251

    Biologist Austin Hughes warns that most inferences of positive selection are based upon questionable statistical analyses of genes:

    A major hindrance to progress has been confusion regarding the role of positive (Darwinian) selection, i.e., natural selection favoring adaptive mutations . In particular, problems have arisen from the widespread use of certain poorly conceived statistical methods to test for positive selection. Thousands of papers are published every year claiming evidence of adaptive evolution on the basis of computational analyses alone, with no evidence whatsoever regarding the phenotypic effects of allegedly adaptive mutations. … Contrary to a widespread impression, natural selection does not leave any unambiguous ‘‘signature’’ on the genome, certainly not one that is still detectable after tens or hundreds of millions of years. To biologists schooled in Neo-Darwinian thought processes, it is virtually axiomatic that any adaptive change must have been fixed as a result of natural selection. But it is important to remember that reality can be more complicated than simplistic textbook scenarios. … In recent years the literature of evolutionary biology has been glutted with extravagant claims of positive selection on the basis of computational analyses alone ... This vast outpouring of pseudo-Darwinian hype has been genuinely harmful to the credibility of evolutionary biology as a science. 19

    Austin L. Hughes, "The origin of adaptive phenotypes," Proceedings of the National Academy of Sciences USA, Vol. 105(36):13193–13194 (Sept. 9, 2008) (internal citations removed).

    Regardless of the above, as I stated earlier the main problem with citing these sources for this issue is that they often automatically assume that human evolved from ape-like creatures over millions of years, to begin with. Hence, while they may be useful within the evolutuionary paradigm, they are essentially useless as evidence against arguments designed to test the validity of the evolution story to begin with.

    I can easily make a program that validate evolution, heck, lets write it right now:
    1) Initialize a population of 1000 individuals, each with a fitness of 1.
    2) (mutation) Add a number uniformly distributed from (-1, 1) to each individuals fitness.
    3) (replication) Delete the 500 individuals with lowest fitness. Copy them to get to 1000 again. Go to 2).
    but that does not prove jack shit because i have not validated it against reality.

    Interesting, since this is actually similar to the operation of the pro-evolution computer simulation that you recommended on this forum (http://www.jehovahs-witness.net/jw/friends/183269/2/Mitochondrial-Eve-for-dummies )

    “The organisms are subjected to rounds of selection and mutation.First, the number of mistakes made by each organism in the populationis determined. Then the half of the population making the leastmistakes is allowed to replicate by having their genomes replace(‘kill’) the ones making more mistakes.” http://nar.oxfordjournals.org/cgi/content/short/28/14/2794

  • inkling
    inkling
    They will dogmatically claim that evolution receives universal support from “all the real scientists” and “all the facts” from genetics, biology, geology, paleontology, and every other “ology” that one can think of.

    How about "scatology"? Becuase that seems to be the branch of science creationists look to most for supporting evidence.

    -inkling

  • bohm
    bohm

    Hooberus: "Actually his program allows for the selection of positive mutations".

    Well im not really sure what you mean, but lets just agree that the statistics and weighting he use on positive vs. negative mutations are stacked so highly AGAINST positive mutations per default that he will allways detect a population crash, though a good mutation may be selected here and there.

    Then you wrote: " Interesting, since this is actually similar to the operation of the pro-evolution computer simulation that you recommended on this forum (http://www.jehovahs-witness.net/jw/friends/183269/2/Mitochondrial-Eve-for-dummies )". My post is not about validating evolution in one swoop. What i talked about is if the mechanisms of evolution can produce information, ie. find non-trivial solutions to hard problems. That is what those programs do, and that is what i argued for (actually it is not an argument, it is a simple observation that has been done over and over again in computer science and engineering).

    This discussion show two things: With some assumptions on how nature works, bad mutations will drown the good and we get a population crash. This is what Dr. sanford show with his computer program. Also we have seen that under another set of assumptions, the principles of evolution can produce novel solutions to hard problems, that is what is seen in GA's and various artificial-life simulations. Can we agree on this?

    Now the big question is what set of assumptions best describe nature? In that case the only thing you can do is to validate your assumptions against nature, so far Dr. Sanford has not done that, and therefore it is not evidence for anything.

    FURTHERMORE you will notice Dr. Sanford make some assumptions that seem quite unphysical, for example tying the positive effect of a mutation to the inverse of the genome length (a-what?), or that he predict humans has only been here for 6000 years (are you really sure you believe in that to?), with the added effect that life forms with the same genome length as humans but a much shorter life span should have gone extinct a long time ago. Unfortunately Dr. Sanford never really address these problem.

  • bohm
    bohm

    Hooberus: And lets have a small talk about scholarship.You can read the following in the last article you cited:

    For example, under Shannon information, which the NCSE would claim is “the sense used by information theorists,” the following two strings contain identical amounts of information:

    String A:
    SHANNONINFORMATIONISAPOORMEASUREOFBIOLOGICALCOMPLEXITY

    String B:
    JLNUKFPDARKSWUVEYTYKARRBVCLTLODOUUMUEVCRLQTSFFWKJDXSOB

    Both String A and String B are composed of exactly 54 characters, and each string has exactly the same amount of Shannon information—about 254 bits. 9 10 Yet clearly String A conveys much more functional information than String B, which was generated using a random character generator.

    Further down he write:

    "In contrast, proponents of intelligent design would define “new” genetic information as a new stretch of DNA which actually performs some different, useful, and new function. For example, consider the following string:

    DUPLICATINGTHISSTRINGDOESNOTGENERATENEWCSI

    This 42-character string has ~197 bits of Shannon information. Now consider the following string longer:

    DUPLICATINGTHISSTRINGDOESNOTGENERATENEWCSIDUPLICATINGTHISSTRINGDOESNOTGENERATENEWCSI

    This procedure just added 42 “new” characters, but no new function has been produced."

    This sound really interesting, and i tried to read about the author, Casey Luskin:

    "Casey Luskin is an attorney with graduate degrees in both science and law. He earned his B.S. and M.S. in Earth Sciences from the University of California, San Diego. ... Casey has published in both law and science journals, including Journal of Church and State; Montana Law Review; Geochemistry, Geophysics, and Geosystems; Hamline Law Review; and Progress in Complexity, Information, and Design. He has published in print and online popular media such as Research News and Opportunities in Science and Theology; Human Events; U.S. News & World Report, BeliefNet; Salvo Magazine; Touchstone Magazine; the Tampa Tribune; the San Diego Union Tribune; the Washington D.C. Examiner, and the Philadelphia Inquire"

    Now something seemed quite odd about his vocabulary, he mix up terms that are not the same but i thought that a man you quoted, who apparently had a degree and has puplished in this field, would know the terms he described surely even a creationist would not make such a stupid-ass mistate to write about shannon entropy without knowing what it is. Boy was i surpriced! The thing about shannon entropy is that you want to define it based on a propability distribution. It is clear from the last quote he calculate it based on the uniform propability distribution and thus arrive at 197 bits for the string:

    "DUPLICATINGTHISSTRINGDOESNOTGENERATENEWCSI"

    However this is completely bogus. If you want to use calculate the information in a signal, you want to use the propability distribution the signal is generated from, ie. you can measure from the signal. Doing that you arrive at something more interesting. First off, in the first example the actual amount of information is the signal is:

    "DUPLICATINGTHISSTRINGDOESNOTGENERATENEWCSI" : 157 bits.

    and for the other Strings:

    A : 211 bits

    B : 232 bits

    As we expect, Shannon entropy allow us to understand which signal is generated from a random number generator, and which is not. This is how it is usually used. And if you want to argue that i should use the uniform distribution, keep in mind that he is talking about detecting design. if we assume no design, ie. completely random numbers, we would expect shannon information to give that answer - i challenge anyone to find a good argument for that use, or any example in genetics where it is used like that. Shannon information is allways used this way i described above when you want to use it as a meaningfull measure in text, image or sound data, except when you are Dr. Casey, apparently.

    Why on earth can a creationist so-called expert in entropy not properly explain what it is, or use it to calculate the entropy of very simple text strings, but still puplish papers about the subject and be seen as an authority?

Share this

Google+
Pinterest
Reddit