Alternate post title: Why Charles Jackson is a tool who can quote papers, but doesnt understand what he is reading.

I get this question all the time, and its totally valid:

How do you tell the difference between an endogenous retrovirus that is shared because of common descent, and a retrovirus that was endogenized independently in two species?

A follow up paper to the one I wrote about here (Me, and you, and Zaboomafoo) provides a lovely example of how we do this!

So heres the back story: There are a ton of different retroviruses. Its not one big homogeneous group of ‘virus kind’, each family of retrovirus has its own genetic features and personality quirks. For a nice comparison so you all can appreciate the diversity of retroviruses: you are as ‘related’ to green algae, as HIV-1 is ‘related’ to a family of ERVs, HERV-K.

Lentiviruses, like HIV-1, apparently dont like to endogenize very much. Weve looked in humans, primates… we cant find any. We found one in a particular species of bunnah. And they just found another one in Cheirogaleus medius, a kind of lemur.

Neato!

Well, another lab looked at different species and genera of lemur… and found another endogenous lentivirus! In another genus of lemurs, Microcebus! Additionally, these lemur ERVs are 93-96% similar to one another, while HIV-1 viruses are only 80-85% similar to one another, max.

At the surface, this might make it look like Microcebus and Cheirogaleus share this ERV because of an endogenization event in a common ancestor– ‘same virus’, two different genera. Right?

Wrong!

You need to remember that where retroviruses insert in a genome is random. Like Ive said a hundred times before, yes, some like to insert near active genes, and some like quiet genes, but exactly where a retrovirus inserts– near which active genes, exactly which nucleotides are up/down stream, is random.

However, there is no complete ‘lemur’ genomic sequence. How can you ‘tell’ where an ERV has inserted if you dont even have a genomic map?

PCR!

To identify putative lentiviral ERVs in lemur genomic sequences, Gilberts lab compared the RELIK sequence to lemur sequences (however little we have). Areas where the sequences matched, they called putative lentiviral ERV sites. Even if you dont have complete lemur genome sequences, you can still see what sequence is directly upstream and downstream from the ERV. In this drawring I made, A is a ‘normal’ sequence, B is the sequence after an ERV has inserted, and C is what happens when an ERV ‘pops out’– the pink parts, the viral LTRs (promoters) line up during cell division, and the gene portions of the ERV just pop out, leaving a ‘solo LTR’ footprint of where the complete ERV used to be.

Because Gilbert knew what sequence was upstream and downstream of the LTR, he could design primers specific for that location, and PCR amplify that region of the lemurs genome. IF there was not and had never been an ERV at that site in the genome, he expected a 250 base pair product. IF there used to be an ERV there, and there was only a solo LTR left (which is what he thought he saw in his RELIK/genome comparisons), he expected a 670 bp product (just an example, he did a few this way).

You can see how different these sizes look in a gel in Figure 5.

He determined that even though the lentiviral ERV sequences were super similar, they are not located in the same genomic regions. ‘Present’ PCR fragment sizes in Microcebus species were ‘absent’ fragment sizes in Cheirogaleus (he also tested this in a similar, but different way, Southern Blots, and other basic molecular genetics tools– Im not explaining them in this post, lol).

He found that with every test, with this particular family of ERVs, there were no orthologous sites of insertion. By using molecular clocks, Gifford determined that though this lentivirus endogenized in two different genera, it happened at about the same time– about 4.2 million years ago. Thus it can be found in the same location in various species of the genus Microcebus, meaning common descent. However it is in a different location in species of the genus Cheirogaleus, meaning Microcebus/Cheirogaleus infections were two independent events.

We can differentiate between common descent and two independent events.

And they didnt even need a complete ‘lemur’ genomic sequence to figure this out.

Comments

  1. #1 MitoScientist
    April 8, 2009

    Whew! Some hot genetics this morning, and I haven’t even had my coffee yet! As always, an excellent post for scientist and layman alike.

  2. #2 Optimus Primate
    April 8, 2009

    Abbie,

    Forgive me if this seems like an ignorant question, but I’m just an interested layman, so I’m plenty ignorant. :)

    You said:

    …they didnt even need a complete ‘lemur’ genomic sequence to figure this out.

    If I’m understanding this post correctly, though, even if both genomes were sequenced entirely, it doesn’t seem to me that this would be any more of a slam dunk. Would it?

    I know it’s a lot of work, but I really think you should animated this and present it in more detail at your next debate. This is kinda awesome. And by “kinda,” I mean “a lot.”

  3. #3 ERV
    April 8, 2009

    If I’m understanding this post correctly, though, even if both genomes were sequenced entirely, it doesn’t seem to me that this would be any more of a slam dunk. Would it?

    Nope!

    You dont need entire genome sequences to figure out ‘where’ an ERV is. You just need to know what your surroundings look like :)

  4. #4 dmso74
    April 8, 2009

    OK, but if ERVs are inserted randomly, does that mean that all the individuals that carry an ERV at the same position are descended from one individual?

  5. #5 William Wallace
    April 8, 2009

    Abbie,

    I am glad you addressed this question, it is one that I have also asked, as you know.

    You need to remember that where retroviruses insert in a genome is random. Like Ive said a hundred times before, yes, some like to insert near active genes, and some like quiet genes, but exactly where a retrovirus inserts– near which active genes, exactly which nucleotides are up/down stream, is random.

    Is this random statement an assumption, or a conclusion well backed up by data?

    How exactly did you or other scientists determine that the insertion points are random? And are you sure?

    Or are random insertion points just a useful model?

    For starters, I’d like to see a representational plot of known versus possible insertion points to see if the data used to form the “random” hypothesis is sparse. If the observations are not sparse, you have other hurdles. But I–somebody outside of your field of expertise–suspect the data is sparse. (You wouldn’t roll a 1000-sided die 3 times and claim evidence that it is a fair die.)

    I am aware that in other disciplines quantization “noise” (error, really) is modeled as a random process, and that this model is useful in many applications, but the fact of the matter is, quantization error is absolutely *not* random. Despite this, and textbook and technical paper authors being very precise in defining the quantization noise as a mere model for quantization error, I have encountered too many university educated individuals who have not paid sufficient attention to this fine point. Quantization error is not random, even if a random mathematical model of quantization error is very often and legitimately used.

    A model’s usefulness has no bearing on whether or not it is true. Other examples of this fact exist in science and engineering. (As another example, Tycho’s sixteenth century geocentric model of the solar system was very useful for constructing navigational charts so that others could accurately sail the world’s oceans. But the model’s utility didn’t make the Earth the center of the solar system.)

    He determined that even though the lentiviral ERV sequences were super similar, they are not located in the same genomic regions.

    Nice to see you’ve corrected this to “super similar”. During the debate, you had stated that they “determined that they were very different retroviruses”.

    But a remaining question is, aside from what I presume is a random insertion point model, how do you differentiate between ERVs currently being at different insertion points, and ERVs that have merely been reshuffled. (In this later case, it is actually to the benefit of the common descent hypothesis. But of course, it is a two edged sword–especially if the random insertion point model is not a reflection of reality.)

  6. #6 W. Kevin Vicklund
    April 8, 2009

    But a remaining question is, aside from what I presume is a random insertion point model, how do you differentiate between ERVs currently being at different insertion points, and ERVs that have merely been reshuffled.

    And Limp Willy proves once again that his reading comprehension is as flaccid as his assets. Did you even read the post?

  7. #7 dmso74
    April 8, 2009

    Kevin,
    I’m no fan of WW, but I don’t think his point, i.e. how to distinguish between ERVs that are in the same spot bc they were inserted there and those that happened to wind up in the same spot through reshuffling, was addressed in the post. I think the answer is simple parsimony, but i didn’t get that from the post.

    my question remains, however: if insertion is random, then the only way an entire species can share the same erv in the exact same spot is if that entire species was descended from a single individual who had the erv inserted in that spot. or am I (likely) missing something?

  8. #8 BAllanJ
    April 8, 2009

    Cool post once again, Abbie! Thanks.

    I was wondering if the fact that these 2 endogenizing events happened about the same time means that the lemurs were having an epidemic at that time. Would there be any other evidence of that? Does anything like this virus still infect lemurs?

    On another track, if I’m understanding this at all, I would think that while the insertion is at a random point, not all points would lead to a viable lemur, so we wouldn’t see them randomly situated in their descendants.

  9. #9 BAllanJ
    April 8, 2009

    Also, if I’m understanding correctly, I have another analogue. Suppose you’re looking at 2 pictures of a house…. they look the same, but you can’t tell if they’re the same house or just 2 houses made to the same plan. You can tell if you can see the houses on each side, or even if you can just tell how far the neighbours houses are from the one in question. You don’t need the actual address (ie complete genome).

  10. #10 Shirakawasuna
    April 8, 2009

    dmso74, that is indeed the inference. For why it’s simultaneously cool and doesn’t violate the principle that populations evolve, check out ‘most recent common ancestor’ on Wikipedia. Imagine a certain protein that all us humans have a simple gene for. At some point, in all likelihood, it arose in a single individual (sometimes relatively common mutations can happen in separate individuals).

  11. #11 Shrunk
    April 8, 2009

    my question remains, however: if insertion is random, then the only way an entire species can share the same erv in the exact same spot is if that entire species was descended from a single individual who had the erv inserted in that spot. or am I (likely) missing something?

    Well, of course, Satan could have inserted the ERV’s after The Fall to make creationists look like idiots. There is that hypothesis. But apart from that: No, common descent is the only reasonable explanation.

  12. #12 Cath@VWXYNot?
    April 8, 2009

    Excellent explanation! I’m sure you’ll be referring people to it often.

  13. #13 dmso74
    April 8, 2009

    Shirak,

    I can see that happening w a beneficial protein, but why would the individual w an erv have such a huge survival/reproductive advantage?

  14. #14 jon
    April 8, 2009

    They don’t need one dmso74…Ever heard of genetic drift?

  15. #15 jon
    April 8, 2009

    or the founder effect for that matter.

    Evolutionary theory isn’t all selection any more. There are many other processes that affect change in populations.

  16. #16 Tyler DiPietro
    April 8, 2009

    “I can see that happening w a beneficial protein, but why would the individual w an erv have such a huge survival/reproductive advantage?”

    They don’t need one. Most of the content of your genome, indeed most genomes, is selectively neutral.

  17. #17 eddie
    April 8, 2009

    “Why would an individual w an erv have a huge survival/reproduction advantage?”

    I think the clue is in the question. The individual would be one of a subset of that species that survived the epidemic and went on to reproduce more that those that didn’t survive.

    Also, I think WW was trying to ask the opposite of dmso74. Namely, how could you tell if two species acquired an erv from a common ancestor and then one of the descendant species shuffled the gene to a different place?
    His point about the validity of random models with the data ret size was uncharacteristically good, but I’m not sure how relevant it is.

  18. #18 cprs
    April 8, 2009

    Authors wrote re the multiple Microcebus pSIVfdl copies, ‘They could result from repeated germline insertions of the same or very similar circulating lentiviruses, intragenomic retrotransposition events, reinfection by an endogenized copy, or a mix of these mechanisms.’ WW has a point which the authors don’t feel able to refute entirely, so why then is Abbie being dogmatic that they result from descent? (as opposed to the well demonstrated and distinct endogenisation in Cheirogaleus). Paradigm blinkers again?

  19. #19 dmso74
    April 8, 2009

    jon and eddie,

    thanks for your thoughts. your two answers work really well together and gave me a duh! moment.. the population gets cut down to just a few individuals thanks to the virus.. then founder effect can take over from there.. could help explain why the speciation event took place as well (founder effect).. neat!

  20. #20 cprs
    April 8, 2009

    One common and well founded answer to the question about advantage of ERVs is protection against XRVs (near identical exoviruses)- well studied in the case of JSRV and eJSRV. (A Paradigm for Virus–Host Coevolution: Sequential Counter-Adaptations between Endogenous and Exogenous Retroviruses Nov 2007. PLOS)

  21. #21 ERV
    April 8, 2009

    cprs/charles– Hi! Glad you made it over to the new ERV. As Im sure you have noticed, your appearance at old ERV was simply bad timing, as I am currently dealing with a Creationist named Charles who doesnt understand ERVs with an obsession with the word ‘paradigm’.

    Unfortunately, second Charles Creationist, you dont appear to understand ERVs any more than the first one.

    Authors wrote re the multiple Microcebus pSIVfdl copies, ‘They could result from repeated germline insertions of the same or very similar circulating lentiviruses, intragenomic retrotransposition events, reinfection by an endogenized copy, or a mix of these mechanisms.’ WW has a point which the authors don’t feel able to refute entirely, so why then is Abbie being dogmatic that they result from descent? (as opposed to the well demonstrated and distinct endogenisation in Cheirogaleus)

    There is more than one copy of this lentiviral ERV in the genus Microcebus. Say there are 30 insertion sites– these 30 could be the ‘result from repeated germline insertions of the same or very similar circulating lentiviruses, intragenomic retrotransposition events, reinfection by an endogenized copy, or a mix of these mechanisms’. The authors of this paper didnt give a shit, because its irrelevant to the goals of this paper– two independent events in two genera, or common descent?

    Lets say there are 30 copies of this ERV due to a mix of all of those listed mechanisms.

    That doesnt change the fact that these insertions are orthologous between Microcebus species. Read the next paragraph of the paper.

    We see this in the human genome HERV-K family. Many members of this family are not independent infectious events. They are the result of a HERV-K reverse transcribing and plopping back into the same genome.

    One common and well founded answer to the question about advantage of ERVs is protection against XRVs (near identical exoviruses)- well studied in the case of JSRV and eJSRV.

    Creationist Claim #6a. Great job!

    Now go read about MMTV for a counter example– endogenous MMTV leads to increased susceptibility to exogenous MMTV, and actually decrease survival rates after infection with some kinds of bacteria.

    Though ERV proteins are occasionally (read: rarely) beneficial, ERVs are not present in genomes because they are beneficial. They are present because of random genetic drift, and because there is no way to get rid of them outside of the recombination event I stated in this post, which still leaves solo LTR footprints.

  22. #22 Strider
    April 8, 2009

    Does everyone who comments here know that the douchebag who commented in the middle of part 7 of the “debate” video is Rhology? I mean, his idiocy in that video is unmatched by any of his comments I’ve ever read here.

  23. #23 Shirakawasuna
    April 9, 2009

    dmso74, your hypothetical pandemic situation isn’t required. Like jon said, think of genetic drift, which can be huuuuuuuge. First, consider that insertions can happen all the time in various places. Let’s consider just those that are neutral, which are going to have a lot of potential insertion areas w/ big free-living eukaryotes like lemurs (introns, etc). Now, the foundering effect would certainly increase the chances of a single mutation being passed on entirely via drift, but it isn’t required, as these things, again, happen very often on an evolutionary timescale. So we look at this one mutation and compare it to the rest of the population: they don’t have the insertion. But, since there’s no selection happening, there’s nothing really stopping it from taking hold in the population and there are many simulations of how this stuff works, essentially taking a drunkard’s walk.

    Allopatric speciations events and all that also make it interesting. Let’s say 12% of the population has this random, non-selected mutation hundreds of years later and *then* a small group of 100 finds its way to a separate location and the groups become isolated. Only it wasn’t a completely even sampling: by random chance, 40% of the founders have this ERV. That’s a foundering effect that had nothing to do with the actual disease but still let this ERV’s frequency in the population go way up!

  24. #24 windy
    April 9, 2009

    WW

    But I–somebody outside of your field of expertise–suspect the data is sparse.

    Oh for fuck’s sake. HOW LONG HAVE YOU BEEN READING THIS BLOG? You don’t recall reading about any other retroviruses that have been found in genomes? Little tidbits like there are about 98,000 ERVs in the human genome alone? Other tidbits like we can *experimentally* map insertion sites?

    If we are talking about lentivirus insertion sites only, there the data indeed is sparser, which was clearly mentioned in the post, and if you had read and understood it you would not need to “suspect”, you cretin!

    So, do you think that because there’s less data on lentiviruses, it’s reasonable to suspect that they, unlike other retroviruses, are somehow are able to home in to the exact same location out of billions of nucleotides in the host genome? Until we find a lentivirus with a little fairy or computer inside it, HOW ABOUT NO!

    I am aware that in other disciplines quantization “noise” (error, really) is modeled as a random process

    What the fuck does quantization error have to do with the distribution of insertion sites in a genome?

  25. #25 Joshua Zelinsky
    April 9, 2009

    William, there’s a fair bit of evidence that ERV insertions will be more or less random (stochastic might be more accurate than random). In particular, retroviruses themselves insert at pretty close to random points. This can be observed in the lab.

  26. #26 Zombie
    April 9, 2009

    Lemurs are cool. That is all.

  27. #27 Jason Dick
    April 9, 2009

    OK, but if ERVs are inserted randomly, does that mean that all the individuals that carry an ERV at the same position are descended from one individual?

    I think the correct statement is that it’s highly likely that they both share a common ancestor that had that particular insertion.

    However, since the insertion points are random, it is possible that, just by chance, two independent insertion events could lead to the exact same insertion of nearly identical viruses in the same location. With all of the tens of thousands of viruses in our genomes and other genomes, it stands to reason that eventually we’ll find one or two such chance insertions.

    Similarly, there is a convoluted sequence of events which can cause a phantom of no common descent where there actually was common descent. That is, imagine if the original common ancestor had a duplicate of the region where the ERV was inserted: one block of DNA with the ERV, one block without. Furthermore, the ERV itself might have inserted itself into more than one location, either originally, or due to copying itself elsewhere in the genome. So, again, this sequence of events is highly unlikely, but just given the large sample size, chances are we’ll see it happen once or twice.

    Therefore, these ERV measures can show to a very high degree of precision whether or not a specific insertion is or is not homologous. But I think more work is required if you’re not content with merely ballpark one-in-a-million odds of being correct.

  28. #28 Jason Dick
    April 9, 2009

    Typo: Oops, meant ballpark one-in-a-million odds of being incorrect. Not the other way around.

  29. #29 Shrunk
    April 9, 2009

    Jason Dick,

    The problem with your argument is that the ERV evidence for common descent is not just that there are common insertions between two species. It’s that there is a pattern of insertions between multiple species, even crossing different genera, that is exactly as predicted by common descent. The odds of such a pattern arising by chance is for all intents and purposes impossible. If what you were saying was true, then we would expect to see some ERV’s in common between humans and, say, chimps, organgutans, gorillas, rodents, etc. But the specific ERV’s would be different between each pair (No reason we would share the same ERV’s with chimps as gorillas, nor any reason that chimps and gorillas would share the same ERV’s that they do with us.) And rodents would share a portion of ERV’s with us in proportion to their genetic similariy to us, not in a pattern that allows the creation of a phylogenetic tree based on ERV evidence alone. The last point is an important one. ERV’s can be used to contruct phylogenetic trees which, by itself, is evidence against coincidental separate insertions. And the capper is that these trees are consistent with those based on independent evidence from, say, the fossil record and other molecular genetic evidence.

  30. #30 dmso74
    April 9, 2009

    SHirak et al.

    My (perhaps mistaken) understanding is that specific ERVs are fixed (everyone has them) in a population at the same spot in the genome, implying that they came from a single insertion event in a single individual. is there any consensus on this issue?

  31. #31 SiMPel MYnd
    April 9, 2009

    Excellent post!! Thanks, Abbie!

    And thanks “Shrunk” for the last comment (#29). The sum-up…

    ERV’s can be used to construct phylogenetic trees which, by itself, is evidence against coincidental separate insertions. And the capper is that these trees are consistent with those based on independent evidence from, say, the fossil record and other molecular genetic evidence.

    …will be so useful in creation smackdown. I’ve been lurking here for a few weeks, and this post and comment are my best take-aways thus far.

  32. #32 William Wallace
    April 9, 2009

    Thanks Joshua Zelinsky. Can you reference a paper (preferably available freely to the general public) that proves this, or even just provides very strong evidence, and doesn’t just assume it?

    Abbie is silent so far, perhaps because she doesn’t want to get pinned down on this point so critical to her argument.

    So, Abbie, do you think that the fact that the Martin et al paper, Retroviral DNA Integration: ASLV, HIV, MLV Show Distinct Target Site Preferences is sufficient to prove your assertion that these are random? Its figure 1 *looks* random, after all. Is “looking random” sufficient to demonstrate randomness in your scientific opinion?

    Or is there another paper you can cite to scientifically justify your assertion, “You need to remember that where retroviruses insert in a genome is random.”?

    Or should we just take it on faith, and not fact check your assertions?

  33. #33 ERV
    April 9, 2009

    Oh my fucking god do you know how to search at ScieneBlogs, WW? Do I need to ask teh management to create a video tutorial to teach idiot Creationists how to use teh Google Machine?

    I addressed that paper a year and a half ago, genio. Way to stay on the ball, there.

  34. #34 William Walalce
    April 9, 2009

    I know you addressed the paper, which is why I brought it up. I am asking you to clarify if you think that a picture with lolipops that appear to be randomly dispersed is sufficient to back up your claim, in your mind, that integration points are random?

    If so, I will spend a bit of time educating you and those who might be misled by your conclusion that random looking data isn’t very good evidence of a stochastic process. (Indeed, the authors of that paper don’t seem to draw the conclusion you do from figure 1.)

    If not, I’d rather you just cite a paper that backs up your claim. Or is the “random” claim, mentioned in your argument, not actually a key point of your argument?

  35. #35 Shrunk
    April 9, 2009

    I’m not sure what point you think you have here, Mr. Wallace. You’re trying to claim that ERV’s keep showing up in the exact same location, not only in every single member of a species but in different species, not because of anything to do with inheritance, but because retroviruses insert into only a few preferred sites in the genome. So all you have to do is cite some evidence that retroviruses do just that. In the paper you are discussing, for instance, there should be many insertions that are identical if you are correct. But, no, every single one of the over 3000 insertions is in a different, unique locus. None of your fancy talk about “quantization error” can hide the fact that can’t point to a single instance of a retrovirus inserting into the same point of a genome on two separate instances. Unless you know of such an instance?

  36. #36 Joshua Zelinsky
    April 9, 2009

    William, why don’t you read Abbie’s earlier discussion of the papers in question or just read the paper you cited. That paper is quite clear that insertion is randomized process with some retroviruses having slight preferences for certain regions.

    Moreover, fixed insertion points wouldn’t give you a nested hierarchy based on ERVs. This should be obvious if you spend about 30 seconds thinking about it.

  37. #37 Matthew Ackerman
    April 9, 2009

    “With all of the tens of thousands of viruses in our genomes and other genomes, it satands to reason that eventually we’ll find one or two such chance insertions.”

    Well, no actually. There are 3,000,000,000 base pairs in the human genome. Therefore the chance that two separate events happen at then same base pair is 1/3,000,000,000 (Pick one base pair at random, what are the chances of picking that base pair again?). Now, if we have 5,000 infections in pairs (so that there are only two of each kind of insertion in the genome), We have only 5,000 chances to hit an event with odds of 1/3,000,000,000 and we have approximately 8 in a 1000 chance of getting one coincidence.

    Now, if we cannot tell any of the retro-viral insertions apart from each other the odds go up very significantly to approximately 4 in 10.

    But there are two points to remember: as Abbie pointed out, we can tell the ERV’s apart from each other, so the actual odds are better represented by the first calculation, and, more importantly, we have not actually checked every one of those 10,000 ERVs, but have only checked the best ERVs, and finally, we haven’t gotten just a single match, but have gotten, what 12?

    Even going off the second calculation I suspect the chances of getting that many hits are very very slight (I can’t think of how to due the calculation just yet, check out http://en.wikipedia.org/wiki/Birthday_paradox for to try it your-self, could you just use a normal approximation?)

  38. #38 William Wallace
    April 9, 2009

    First, the important question, and the only claim I have made so far, is that the random appearance of data doesn’t establish that it is random. It could be random, but, especially in the case of sparse data, there could be latent relationship.

    I will spend some time illustrating the shaky ground she’s on, if Abbie confirms that the random appearance illustrated in figure 1 is the basis of her claim.

    Nice goal post shift attempt, Skunk.

    We’re talking about Abbie’s claim. “We need to remember…”.

    By “remember” does she mean “take on faith” or does she mean that there is a study that establishes once and for all that the integration points are stochastic?

    Joshua, I did read the discussion. Where do you think I found the paper. I’ll quote Abbie:

    Look at Figure 1: All those blue lollipops are places they found where HIV inserted itself.[Figure 1 ommitted]…Theres a lot more than one lollipop in that figure.

    Is this the best evidence she has? Appearance in a graph? Note she has still not confirmed it, because, I suspect, she knows this straw man of data appearing random as strong evidence of a stochastic process is easily slain as very weak evidence.

    Joshua, the opposite of random isn’t “easily predicted”, or even “fixed”. I’m not pointing this out for your sake, because I suspect you know this. But for the sake of others, and for the sake of clarity.

    That paper is quite clear that insertion is randomized process with some retroviruses having slight preferences for certain regions.

    My understanding is that comparison datasets were selected randomly to remove bias, but my reading of the paper is that it concludes that a bias was detected. This bias was not completely characterized, just reported. The bias rules out a rectangular PDF, but doesn’t authoritatively characterize the distribution of ERV integration, does it?

    And even if it did, the analysis of a mere 3,127 integration sites is not prima facie evidence that ERV insertion is a stochastic process. Indeed, I didn’t find in the paper a description of the criteria used to select the 3,127 integration sites. Was every known integration site examined? Were duplicates removed? Could duplicates even be identified? Are the 3,127 integration sites claimed to be ergodic?

    So, once again, slaves to evidence though some of us claim to be, the request for evidence that ERV insertion points are random remains glaringly unanswered. If the insertion points are random, so be it. I don’t know. But convince me scientifically, don’t just tell me that I “need to remember”. I am not a chance worshiper, who falls for the “stochastic process of the gaps” argument, but neither do I deny that chance happens.

    The question is, are ERV integration points stochastic, and what paper or papers, preferably publicly available, have settled this question once and for all?

    Assumptions aren’t bad, as one of Abbie’s least favorite teachers recently pointed out, but you should be aware of them, disclose them, and be aware of the risks involved in basing conclusions on them.

  39. #39 Matthew Ackerman
    April 9, 2009

    Shrunk: “So all you have to do is cite some evidence that retroviruses do just that. In the paper you are discussing, for instance, there should be many insertions that are identical if you
    are correct.”

    As a general policy I would like people to simply ignore W.W. since it is highly improbable that he will say anything important, but I will use the shenanigans to launch into scientific philosophy.

    You make an excellent point and I would like to make it explicit why the burden of proof is on W.W. to show that retro viral insertion is not random, and not on Abbie to show that it is.

    Proving that something is random is essentially impossible, and to attempt to do so is non-scientific*. Why? Saying that something is random is equivalent to making the claim ‘there is no pattern which can explain these data.’ A moments reflection will tell you why this cannot be proven. The assumption that data has no pattern is generally the null hypothesis. We simply assume that the data is random until we have reason to think otherwise.

    A passing familiarity with cryptography will serve to illustrate the point. It is very easy to see that a stream of data contains an encrypted message once the message has been decoded (For instance do you believe that this post is just a steam of random letters that just happened to cary meaning by chance?) however, it is actually impossible to be sure that a stream of data does not contain an encrypted message. A ‘one time pad’ encryption is impossible to crack http://en.wikipedia.org/wiki/One-time_pad, so you can never be sure that a message wasn’t encrypted using this method.

    I could go on to address W.W. nonsense specifically, however, the broader philosophical issue was the only part that interested me.

    Just my two statistically significant cents.

    *This isn’t absolutely true, just a useful heuristic.

  40. #40 eyesoars
    April 9, 2009

    WW accidentally impinges on something slightly relevant to the topic that he’s trying to address, and that regards the actual randomness of inserted ERVs.

    The positions that we see historic ERV insertions in segments of an animal population are not going to be perfectly random. There will undoubtedly be slight biases due to the accessibility of the virus to particular partions of the genome. So it’s almost certainly not a perfectly uniform distribution that we’ll see ERV insertion; it will be biased by the infected cells’ use of their DNA and probably the particular methods the ERVs use to insert themselves.

    Secondly, there will also be a substantial bias away from various parts of the genome: insertion into any critical part of the genome will prevent the organism or an infected germ cell from producing viable offspring.

    So, yes there are biases in the location of ERVs in the genomes. No, they don’t affect the validity or results of the study — there are still huge numbers of potential insertion sites.

  41. #41 William Wallace
    April 9, 2009

    While Matthew Akerman is correct when it comes to analysis of data encrypted with a one-time pad, anybody familiar with Johnson noise, Brownian motion, or quantum mechanics can tell he is engaging in sophistry. Consider this: one-time pads cannot exist without a source of randomness. Somebody get a fire extinguisher in here–another straw man just went up in flames.

  42. #42 Shirakawasuna
    April 9, 2009

    dmso74, you got it. Even with reshuffling, the chances of it reshuffling to that *exact* spot on the genome are pretty low, so it’s a good inference that common ancestry is responsible for those shared ERVs.

  43. #43 Shirakawasuna
    April 9, 2009

    WW, even if there’s a latent pattern (which you are concocting in your head), the data points to the insertions being all over the place, indicating a large amount of effective randomness. Take your own example of Brownian motion: there’s a lot of randomness in there, even if trends can be identified, and it is a stochastic process. For something even more complicated, there’s random walks, which again illustrate this randomness.

    No one asserts that they have absolute knowledge of a process being random, and that’s what someone would need to avoid your silly tangential claim: ‘it could still have a trend of explanation!’. Instead, we look at the actual data and see what patterns there are. So far, it sure looks like insertions can happen all over the place and unless you’d like to point out a strong correlation, you can STFU and go read a book.

  44. #44 Shrunk
    April 10, 2009

    Thanks for the explanation, Matthew Akermann. I agree, it’s pointless to expect anything but nonsense from an evolution-denier. Still, it’s fun to push them to see just how ridiculous their arguments will become.

    If I understand you correctly, we can never say a pattern is random because we can never be certain that a non-random pattern exists that would be evident if we had a larger data sample. For instance, suppose we have a series of a million digits that continually repeat in the same pattern, but otherwise have no fixed relationship between each other. If we were to analyse those digits over two or three repetitions of the cycle, we would be able to discern the pattern. However, if we stopped anwhere before the millionth digit, the series would appear completely random. Is that kind of that you’re saying?

    In this specific example, since AFAIK we have no documented examples of retroviruses independently inserting into identical sites, there is no possibility of determining that there is preference to particular loci. However, and this is where I begin to get way out of my depth, it may be possible to determine what kind of selectivity is necessary to support Mr. Wallace’s belief. Remember, what he is claiming is that retroviruses are so picky about where they insert that ERV’s have independently arisen in the exact same location, down to the precise base pair, in every single human, chimpanzee, gorilla. orangutan, and many species of monkey. And that this is not due to inheritance, but to a large number of identical insertions. Even the ERV’s fixed within a single species would not have resulted from common descent (if you are denying common descent as an explanation for interspecific common ERV’s, then I don’t see why you would accept it for intraspecific ones.)

    Now it seems to me it should be possible to estimate what degree of specificity of retroviral insertions we should observe in order to produce this result. Mr. Wallace says 3000 insertions is too small a sample, but I wonder how he knows this? He says he bases his claims on evidence, but has yet to produce any for his “picky retrovirus” hypothesis. My gut feeling is that this degree of specificity should be readily observable in real time i.e. we should be finding retroviruses uncannily always appearing at the same limited number of points in the genome when we analyze new infections. (It’s worth noting such a finding would be of great practical importance, as it would provide a potential mechanism for precisely directing genetic therapies.) But as I say, that’s just intuition on my part. Surely Mr. Wallace or some other “creation scientist” would be able to mathematically determine the degree of specificity we should be observing. Of course, this would only be able to falsify, and not confirm his hypothesis (i.e. if it turns out that, in the number of insertions we have documented, identical insertions should have been observed, and they haven’t been, then the hypothesis is disproven. If it turns out we don’t yet have enough observed insertions to rule out the hypothesis, it still doesn’t mean it is true). However, since creationists claim to be real scientists, that shouldn’t deter them.

    How about it, William Wallace?

  45. #45 Matthew Ackerman
    April 10, 2009

    “If I understand you correctly, we can never say a pattern is random because we can never be certain that a non-random pattern exists that would be evident if we had a larger data sample.”

    Actually, that is not what I am saying. I am saying: When we say that a pattern is random, we are under no obligation to provide rigorous scientific proof of this assertion. If someone else believes the pattern to be non-random, then they must demonstrate how it is non-random. A belief that it is random is our default position until we have some reason to think otherwise.

    As for the second part of your post, it is part of a discussion with W.W., and consequently I will have to ignore it. ;)

  46. #46 Albatrossity
    April 10, 2009

    I don’t know why you are allowing WW to derail this thread into a conversation about whether or not the insertions are random. That may be an interesting mathematical discussion, but the evolutionary significance of these insertions is in relation to common descent. The SITE of the insertions in the genome can be used to detect the difference between insertions inherited from a common ancestor and independent insertions in different organisms in a lineage.

    Common descent is the only rational explanation for the many insertions that are found in the genomes of humans and chimps, since the sites of the insertions are identical. It doesn’t matter if they are random or not; the identity of the sites is the important parameter. The work discussed by Abbie above removes another weak apologetic argument (i.e., maybe they were inserted independently in chimps and humans, how would you know if that was the case?). We can tell the difference. And chimps and humans still have a common ancestor, one proof of which is the number of insertions at common sites in the genome.

    So, WW, how about paying attention to the real issue? What explanation do you have now for the common insertions at homologous sites in the genomes of chimps and humans?

  47. #47 Shrunk
    April 10, 2009

    Matthew (#45):

    I think we’re really saying the same thing. My point is that to say something is non-random is an unfalsifiable position. If there is no evidence of a non-random pattern, the creationists could always say, “You need a larger sample” and they could continue to do this ad infinitum. At some point, however, it is possible to say, “Even if we did find some evidence for preference of particular sites by using a larger sample, at this point we can say that any such preference must be too weak to explain universal fixation of these ERV’s without resorting to common descent as an explanation.” My question to WW (you can stop reading here, Matthew) is why don’t creationists try to define the sample size at which that statement can be made, thereby formulating a falsifiable hypothesis that would allow creationism to be dismissed, by creationists’ own criteria, once and for all? That’s what any real scientist would endeavour to do.

  48. #48 jon
    April 10, 2009

    @Shrunk:

    Because they like it that way. They simply can’t win without the use of a grab bag of logical fallacies.

    This randomness thing looks like a mixture of moving the goalposts and pointing at perceived “gaps” to me. No matter how large the sample size gets, they’ll always say, “well it just isn’t big enough yet!” and deny the research. Then they’ll claim that this is evidence for the situation being non-random when, in fact, it isn’t at all.

  49. #49 cprs
    April 10, 2009

    #21 Sorry to reply so late, but I’m well out of sync with OK and busy. Thanks too for your patience in answering someone you clearly regard as troublesome.

    I don’t profess expertise in ERVs, but I do profess fascination, and I envy your opportunity to study something so intriguing – ERVs necessitate major paradigm shifts from classical genetics and standard epigenetics whatever one’s prior assumptions, especially given the intricate functional ties between ERV genes and LTRs and host genes and inducers.

    The question you posed here, Abbie, is seminal, ‘How do you tell the difference between an ERV that is shared because of common descent, and one endogenized independently in two species?’ By claiming all the Microcebus insertions are orthologues either you are presuming what the authors do not or you’re using a different definition of orthologous than I am used to (homologous genes from a common ancestor). My assertion stands – are you not being dogmatic without proper grounds? I am not at all convinced that the paper answers your question neatly – not yet anyway.

    Who claims defence against XRVs, specifically JLR (JSRV late restriction), is universal? What’s interesting is that the inactivating mutation is apparently not random – so suggest the authors – (‘the R21W substitution is not strictly necessary for JLR, as single JSRV mutants carrying amino acid residues other than R at position 21 are incompetent for viral release and transdominant over JSRV’)- they also hint at host exploitation of endogenisation (http://jvi.asm.org/cgi/reprint/81/4/1762) – the MMTV counterexample is interesting but it’s not so surprising to see ERVs facilitating as opposed to blocking XRVs.

    By the way, when you claim ERVs are rarely beneficial what do you make of the Atlanta group’s claims that 51,197 ERV-derived promoters intiate transcription in man, including 114 start sites which are responsible for 97 genes (Bioinformatics 2008.21.12.1563-67)?

    Best wishes, keep up the blog – even though your subtitle is blasphemous and does you no credit. I’ll be following even if I don’t reply again.

  50. #50 cprs
    April 10, 2009

    Orthologous applies to host genome position not ERV origin, I understand better, and should have read your reply more carefully.

  51. #51 William Wallace
    April 10, 2009

    Shrunk, you’re a very dishonest person. Your claimed summary of my arguments is not an accurate reflection of my arguments. Matthew Ackerman has already pointed out that you did not summarize his argument accurately, either, so you’re getting it from both sides.

    To address the point about randomness, you can see in #5 that I am asking for evidence, good evidence, not proof. Evidence of randomness can come in a variety of forms, from weak, to strong.

    The reason I am asking for this is because Abbie based her argument on the random assumption. If you look at #38, I said it is okay to have assumptions, but the argument should be properly stated. What Abbie is really arguing, it seems to me, and you’re welcome to correct this if I am wrong, is ~”Based on the assumption and preliminary data that ERV integration loci is random variable, and the assumption that, once integrated, ERV sequences do not get reshuffled intact, we can differentiate between quasi-contemporaneous ERV integrations in different species from ERV integrations that were inherited from a common ancestor infected once.”

    For instance, suppose we have a series of a million digits that continually repeat in the same pattern, but otherwise have no fixed relationship between each other. If we were to analyse those digits over two or three repetitions of the cycle, we would be able to discern the pattern. However, if we stopped anywhere before the millionth digit, the series would appear completely random.

    Shrunk, this example is valid, with some additional clarification that I won’t delve into. There are interesting counter examples, as well. If you consider quantization error modeled as noise, for example–if you analyze an ergodic ensemble of real valued continuous time signals against their quantized discrete time counterparts, you will see that quantization errors have the characteristics of random noise. However, if you take a specific instance of a real time signal compared to its sampled and quantized counterpart, and analyze its quantization error, it will not be noisy at all. The quantization error will not only be deterministic, it will be periodic for a periodic input signal. So, in this counter example, less data–in the way of a single signal–shows a readily discernible and deterministic pattern. Whereas more data–in the way of an ensemble of ergodic signals–shows a pattern of quantization error having the characteristics of a random variable, hence the useful quantization noise model.

    [#46]I don’t know why you are allowing WW to derail this thread into a conversation about whether or not the insertions are random.

    Simple. It is not derailing. Abbie in the OP said that we need to remember that insertion locations are random. The point is critical to her argument, and stated by Abbie as though the matter is settled.

    What explanation do you have now for the common insertions at homologous sites in the genomes of chimps and humans?

    Are you suggesting in effect that only a restaurant owner can be a restaurant critic?

    I’m just trying to evaluate the strength and weakness of Abbie’s argument.

  52. #52 Joshua Zelinsky
    April 10, 2009

    William whether the process is random, statistically random, or stochastic doesn’t make any difference. The analysis functions fine.

    We can see modern retroviruses (such as HIV) insert themsleves into many different spots so we know that retroviral insertion. So there’s strong evidence that the insertion positions are more or less statistically random.

    And again there’s the basic issue you seem to be missing: If there were ordered, highly preferred insertion points then we wouldn’t see a nested hierarchy. Saying something like “well maybe the ERVs insertions are non-random” doesn’t even deal with the matter.

  53. #53 LanceR, JSG
    April 10, 2009

    Limp Willy: the master of the “Have you quit beating your wife yet” question. He does nothing but lie, obfuscate, and ignore any evidence that contradicts his narrow worldview. Then he somehow pretends that he is “just asking”.

    How about it, Limp Willy? Quit beating your wife yet?

  54. #54 Shrunk
    April 10, 2009

    To WW,

    I don’t know how I’ve been dishonest in simply reiterating the fact that it’s not up to us to prove that retroviral insertion is random. Again, simply providing an example of two independent yet identical retroiral insertions is what’s required for you to at least have a basis for starting the discussion. That by itself won’t be evidence for non-random insertion, of course. But without that you don’t even have the barest basis to even start considering this as a valid hypothesis, no matter how much impressive-sounding gobbledygook you can manage to type about “quantization error” and “ergodic ensemble”. Are we supposed to be impressed by that? Sorry, not.

  55. #55 Albatrossity
    April 10, 2009

    When I asked What explanation do you have now for the common insertions at homologous sites in the genomes of chimps and humans?

    WW weaseled: Are you suggesting in effect that only a restaurant owner can be a restaurant critic?

    No. I’m suggesting that by concentrating solely on a definition of randomness, you are avoiding the critical point made by Abbie in the OP. And, unsurprisingly, you are still avoiding it. So I’ll ask again, politely.

    What explanation do you have now for the common insertions at homologous sites in the genomes of chimps and humans?

  56. #56 Shirakawasuna
    April 10, 2009

    It’s very simple, William Wallace. So simple that you should be lobbing those accusations of dishonesty at yourself.

    WW said:
    “To address the point about randomness, you can see in #5 that I am asking for evidence, good evidence, not proof.”

    A distribution with relatively similar frequencies throughout the gnome. You know, like what we have. Like others have said before, however, the null hypothesis is that there isn’t a general pattern: that is, you have you posit and show that one exists, which is reasonable given even diffuse data. Additionally, if new data coming in showed a correlation, you’d *still* have to deal with the data we have showing that ERVs exist and are inserted into relatively inactive areas.

  57. #57 Shrunk
    April 10, 2009

    I hope this won’t be redundant to everyone here, but I thought I’d offer this explanation for why the observed pattern of ERV insertions is considered irrefutable evidence of common descent.

    Let’s imagine we have a board containing 3 billion slots representing the human genome (each slot a base pair). We also have a bag containing 90,000 marbles, each representing one of the ERV’s contained in that genome. Each time we pull a marble out of the bag, we then reach into another bag that contains 50 million different marbles, each denoting a specific slot on the genome-board (50 million being the minimum number of available sites available for a retrovirus to insert). Once we have pulled a marble out of the second bag, we then but the first (ERV) marble into the appropriate slot on the board.

    Now lets suppose we perform this task with a human, a gorilla, a orangutan and a chimpanzee, each pulling marbles and placing them on the board in the procedure described above. Obviously, the task will take a rather long time, but at the end we look at each primate’s board to see how many marbles have been placed in the same slots. Now suppose you find that each board has 12 of the marbles in exactly the same place, and there are no other marbles in common between any two boards. Would you be surprised? So surprised that you would in fact be certain that something else must have explained this besides random chance?

    There’s more. Because ERV’s contain a matching pair of LTR sequences which are subject to random mutation, an ERV can function as a molecular clock to allow estimation of the time of insertion. As it turns out, this method confirms that the common ERV’s have been inserted at the same time. So, in our game, this would mean that those twelve marbles were not only placed in the same slot for all for primates, but they were drawn from the bag at the same time in all four instances. I’ll let the more mathematically inclined calculate the odds of this happening by chance, but I strongly suspect it is somewhere between, “Yeah, right!” and “As if!”

    Yet even this is an underestimate of the unlikelihood of what we actually observe. It’s easy to forget the reason such a big deal has been made over ERV’s. It’s true that they provide strong evidence for common descent, but everyone who’s not a faith-addled creotard already knew common descent was real. What was exciting about the finding was that they provided a tool to create phylogenetic trees as a means of confirming or clarifiying phylogenetic relationships based on other paleontological and molecular evidence. So even my analogy, complicated as it is, is oversimplified. A more accurate one would be if we had predicted a specific pattern of common ERV marbles (more in common between chimps and humans than between humans and the other two) and ended up with that exact pattern.

    Hope that helps more than it confuses….

  58. #58 James F
    April 10, 2009

    D’ja ever notice how creationists invoke probability when arguing against evolution in general, but then when faced with ERVs the factor of probability goes right out the window? Keep on keepin’ on, guys!

  59. #59 Optimus Primate
    April 10, 2009

    Shrunk, you’re a very dishonest person.

    Posted by: William Wallace | April 10, 2009 12:31 PM

    IRONY was found dead on the popular science blog ERV on the afternoon of April 10, 2009. Funeral services will be held on Saturday, April 11, 2009 at 2:00 p.m. The bereaved ask that you send donations to NCSE in lieu of flowers.

  60. #60 386sx
    April 10, 2009

    D’ja ever notice how creationists invoke probability when arguing against evolution in general, but then when faced with ERVs the factor of probability goes right out the window? Keep on keepin’ on, guys!

    That’s because they have bigger probabilities. They take their probabilities and, like, multiply them together and stuff. You ain’t seen probabilities until you’ve seen a creationist probability.

  61. #61 LanceR, JSG
    April 11, 2009

    And like the lying weasel he is, Limp Willy *always* runs away when confronted with his lies.

    Come on, big man. Answer the damn questions! You’re so smart and edumacated, and you can’t answer simple questions? One might almost start to consider that you don’t really know anything at all!

  62. #62 William Wallace
    April 11, 2009

    I suppose you believe the retort that any phenomena not fully understood can be attributed to chance, and the burden of proof is *not* on the one asserting that it is chance, but on the skeptic seeking a scientific justification for the assertion.

    If you believe the forgoing is a fair representation of the argument you believe has not been addressed, cut and paste the preceding paragraph into vi (I understand there are some software engineers here), and then issue the command

    :%s/chance/God/g

    Then, read again, let me know what you think.

    And note, “slave to evidence” Abbie has not yet addressed the point first brought up in [#5].

  63. #63 cprs
    April 11, 2009

    Some of these ad hominems against WW are themselves very damning for their authors: they reveal intellectual impotence, immaturity or both. There are real issues here let them be addressed.

  64. #64 Lee
    April 11, 2009

    cprs – they are not real issues – this is garbage willie keeps throwing against the wall. Then, when it doesnt stick, willie keeps pointing to the spot it fell from and saying, why wont y’all tell me what keeps that garbage there on the wall. At a certain point, it gets tiresome to keep pointing out that the crap didn’t stick, and there is no garbage on the wall.

    The issues have been dealt with, ad nauseum. His central issue has been dealt with here, quite well. Willie keeps hammering it nonetheless.

    People insult him, because they are tired of responding to his crap over and over and over and over and over and over and over and over and over…..

  65. #65 Optimus Primate
    April 11, 2009

    @cprs #63:

    On the contrary, they simply reveal that we’re sick of WW’s constant obfuscation, snotty attitude, and deplorable reading comprehension skills.

    He’s like a child who has learned a few big words and gets a kick out of playing the “I’m not touching you I’m not touching you I’m not touching you I’m not touching you I’m not touching you I’m not touching you” game.

    It’s like trying to have an honest conversation with a first-year Philosophy student: boring, infuriating, and ultimately pointless.

    WW has demonstrated no reasonable desire to have an actual conversation: he seems interested only in throwing what wrenches he can into the machinery and claiming they slipped when caught red-handed.

  66. #66 Shrunk
    April 11, 2009

    In the digression caused by WW’s desparate “non-randomness” argument, I think an important point has been lost: Charles Jackson presented the lemur study as a refutation of the ERV evidence for common descent, but there is no way anyone who understands the paper could make that claim. So the options are 1) Jackson cites papers in his talks that he does not even understand; 2) Jackson does not even understand the ERV evidence for common descent even though he claims to be competent to debate on the topic of molecular evidence for evolution; 3) Jackson does understand the paper and the evidence, but intentionally misrepresents them on the assumption that his audience will be sufficiently ignorant not to realize he is doing so. If (3), then it is clear his assumption is correct, for the creationist part of the audience, that is; the video clip linked in Abbie’s post was actually put up by a creationist, oblivious to the fact that this was a major gaffe by their man.

    If you’ve read the comments on the debate itself, you’ll know that I sent Jackson an email to clarify whether he actually understands the ERV evidence, but all he responded with was a request for another debate. I’m not sure if he thought I was connected with Abbie in some way. Weird.

  67. #67 Shirakawasuna
    April 11, 2009

    cprs: “Some of these ad hominems against WW are themselves very damning for their authors: they reveal intellectual impotence, immaturity or both. There are real issues here let them be addressed.”

    If your idea of ‘intellectual impotence’ and immaturity is using insulting language with a trolling, lying idiot, I think you have some growing up to do yourself. If you don’t like negative verbiage on the internet, you can ignore it like mature individuals.

    If you’re referring to ad-hom as an actual fallacy, then it’s again to be expected when the individual is *actually* a trolling, lying idiot. You’ll notice he’s largely ignored the direct responses to his inanity. This is because he’s also a coward. I have to wonder how long you would entertain someone so obviously dishonest. Presumably for weeks, given your clear moral superiority.

  68. #68 William Wallace
    April 12, 2009

    Shirakawasuna’s challenge: Shirakawasuna, why don’t you document some of my supposed lies.

    Optimus Primate and Lee, Abbie made an assertion, but she hasn’t offered a citation to a paper published in a peer reviewed scientific journal that concludes, and not just assumes, that the insertion points are random. I had to search on my own, and the best I could find was her claiming that integration points are random because figure 1 of some paper (that shows a bias) has no pattern she can discern.

    Other than variants of “yes this is true, erv insertions are random” I haven’t ignoreed much on this thread at all.

    Indeed, I am dammned if I do, dammned if I don’t. If I respond, I am a troll, but if I don’t, I am a coward [#61].

    It seems that the trolls here are the ones who are not addressing the issue at hand, but instead resorting to the types of responses you’d expect to see on a Jerry Springer show. Give a journal citation, show the evidence, reframe the argument, or admit that it is based on an assumption without strong scientific evidence.

  69. #69 Shirakawasuna
    April 12, 2009

    “#68
    Shirakawasuna’s challenge: Shirakawasuna, why don’t you document some of my supposed lies.”

    How about you take my *actual* challenge and reply to the many and various points that refute your whining?

    By not doing so and acting like you have, you prove the dishonesty.

  70. #70 Shrunk
    April 12, 2009

    WW: I had to search on my own, and the best I could find was her claiming that integration points are random because figure 1 of some paper (that shows a bias) has no pattern she can discern.

    And in this search of yours, did you uncover any evidence for non-random insertion that we’ve been asking you for since this thread began, and which you have yet to provide?

    Perhaps you could explain how we would discern preference for a specific site when we have 3127 retroviral insertions with no duplication between any of them. Do we give each virus a survey form to rate how much they like the particular site they landed in? If you do decide to go against form and actually answer this question, please do so with relying on obscurantist jargon.

    On a side note, it’s interesting to note that WW is essentially admitting that his belief in creationism now rests on the barest thread: The hope that retroviruses possess some mysterious ability to guide themselves to specific points in the genome. No evidence for this, of course, but he’s convinced that it must exist somewhere. He has no choice, really. The only alternative is to accept the reality of common descent.

  71. #71 Albatrossity
    April 12, 2009

    WW speaks the truth: It seems that the trolls here are the ones who are not addressing the issue at hand

    Unfortunately he is the troll, under this definition and many others. Twice now he has ignored this question, which is central to the points raised in the OP. Will WW whiff again on the third attempt? Let’s find out.

    WW, what explanation do you have now for the common insertions at homologous sites in the genomes of chimps and humans?

  72. #72 Brian X
    April 12, 2009

    WW:

    Here’s the thing: why shouldn’t insertion points be random? Viruses are like bad dates — they shoot and leave. They have no mechanism to make sure where in the genome their DNA actually winds up; barring an assertion of some kind of cellular channel that specifically channels foreign DNA to a specific dumping spot (something that I don’t think there’s any evidence for), the burden of proof is on someone who asserts that it isn’t random.

    Frankly, you sound like Dilbert’s boss being outraged at finding out that 40% of all sick days are taken on Monday and Friday.

  73. #73 Stephen Wells
    April 12, 2009

    It’s even worse for little Willy. All we need is _one_ case where the same retrovirus inserts in two different places- and the paper Willie’s trying to argue over lists thousands- and Willie can’t claim that all the identical viral insertion points are not evidence of common descent.

  74. #74 William Wallace
    April 12, 2009

    [#72], that’s a good line. But see [#62]
    [#71], see [#51].
    [#70], see [#5] (dice example).
    [#69] see above.

    And for the PWND ERV fanboys present at the Jackson/Smith debate, see Evolanders PWND in creationism versus evolutionism debate.

  75. #75 dmso74
    April 12, 2009

    Let’s assume that ERV insertion really is random. then the logical inference is that ERVs shared by common descent must have come from an endogenization event in ONE individual. then all, for example, chimps and humans are descendents of that one individual ancestor. and this must hold true for all ERVs shared at the same genomic location in different species. I am no expert, but it seems like these shared ERVs are pretty numerous, meaning that all the species that share them also came from one individual. does this not strike anyone else as a bit hard to swallow?

  76. #76 Stephen Wells
    April 13, 2009

    @75: Actually, if you think it through, you’ll see that all members of a species have a common ancestor if you go back far enough; read up on a little population genetics. For fun, get to grips with why the last common female ancestor of the whole current human race never met the last common male ancestor… Your problem is that you’re thinking “came from one individual” in an Adam-and-Eve sense, when in fact the “last common ancestor” was a member of a population.

    @74: my @73 kills your attempt at an argument and you know it. And one day you’ll have to explain what the fuck an evolander is supposed to be.

  77. #77 Rrr
    April 13, 2009

    @76: Isnt that a suv-species? ;-)
    @75: Im not a scientist either, but I think you are on to the right line of reasoning here. Its just that final, hesitant straw-clutching sentence you need to get over and beyond.
    If this simple, parsimonious explanation fits all available evidence, I personally find it a lot harder to swallow a goddiditinanonknowablewayevar model, which does not.

  78. #78 Albatrossity
    April 13, 2009

    WW pretends to have answered my question (WW, what explanation do you have now for the common insertions at homologous sites in the genomes of chimps and humans?) in comment #71 with this: [#71], see [#51].

    Unfortunately for WW, his comment #51 has no direct answer to the question other than a snide and irrelevant retort about restaurant owners and restaurant critics.

    So I think that the record shows, per usual, that WW can provide nothing except snark and diversions in support of his notions. He can’t answer questions, but constantly pretends to have done so in the hope that nobody notices his empty suit.

  79. #79 jon
    April 13, 2009

    @dmso74

    It’s not hard to swallow at all…I really don’t get why you don’t understand that while these endogenization events, like mutations in general, occur in individuals and then take hold in the population due to drift, selection, or some other mechanism.

    Really, I thought we went over this at the beginning of the thread.

  80. #80 cprs
    April 13, 2009

    For my sake, if not also for WW’s, please clarify the evidence that germ cell ERV inserts are actually random (preferably proviruses not just LTRs).

    Of the two papers Abbie cites as ‘abused’ by creationists (claim#2), one refers to HeLa cells (immortalised cervical cancer cells) and the exoviruses do show highly preferred insertion sites (as Abbie rightly points out) – HIV for example targetting active genes (presumably at unfolded DNA sites). This does not seem a solid place for data for obvious reasons. The other paper reports one and alludes to two other examples of independent same site insertion.

    Can I also ask clarification for why the CERV 1 family is not found in both man and orangutans, but is in other ‘closer’ primates (I recognise only 4% of these insertions are thought to be orthologous)? (http://biology.plosjournals.org/perlserv/?request=get-document&doi=10.1371/journal.pbio.0030110)

    The same question applies to the CERV 2 family not found in man, orangutan, but it is found in orthologous sites in ‘closer primates’ and macaques (supposedly an older common ancestor, fig.5)(http://genomebiology.com/2006/7/6/R51). The authors also write, ‘we were able to detect pre-integration sites at those regions in the human genome orthologous to the CERV 2 insertion sites in chimpanzees, effectively eliminating the possibility that the elements were once present in humans but subsequently excised.’

    Finally isn’t the site for the recent CERV30 insertion (so called HERVK 10) orthologous between chimps and humans (1p31)?

    Doesn’t this make arguments for common descent based on orthologous insertion sites less likely, unless the standard trees of descent are mistaken?

  81. #81 Shrunk
    April 13, 2009

    dmso74 (#75)

    I’m not sure if what’s giving you trouble is that you’re assuming the first individual who developed the ERV is also the most recent common ancestor of the population in question. If so, you’re mistaken. The ERV was most likely endogenized by an individual who lived some time before the MRCA and that ERV would have existed for a time in a proportion of the population but not every single individual in that population. Eventually, however, most likely just by chance, that ERV became fixed in the entire population and could be traced back to the MRCA.

    Think of it this way: If there is a particular allele that you share with every single one of your cousins on your father’s side then one of your paternal grandparents must have also had that allele. That grandparent is also the MRCA of you and your cousins. However, that grandparent was most likely not the first person to possess that gene.

    This is just basic population genetics, BTW, and has little directly to do with evolution. Even if you’re a creationist, you still have to accept this.

  82. #82 Shrunk
    April 13, 2009

    dmso74 (#75)

    I’m not sure if what’s giving you trouble is that you’re assuming the first individual who developed the ERV is also the most recent common ancestor of the population in question. If so, you’re mistaken. The ERV was most likely endogenized by an individual who lived some time before the MRCA and that ERV would have existed for a time in a proportion of the population but not every single individual in that population. Eventually, however, most likely just by chance, that ERV became fixed in the entire population and could be traced back to the MRCA.

    Think of it this way: If there is a particular allele that you share with every single one of your cousins on your father’s side then one of your paternal grandparents must have also had that allele. That grandparent is also the MRCA of you and your cousins. However, that grandparent was most likely not the first person to possess that gene.

    This is just basic population genetics, BTW, and has little directly to do with evolution. Even a creationist has to accept this.

  83. #83 Shrunk
    April 13, 2009

    Sorry for the double post.

  84. #84 Shirakawasuna
    April 13, 2009

    dmso74… I think I recommended in my first post that you check out what a most recent common ancestor is. It should explain everything.

  85. #85 William Wallace
    April 13, 2009

    Re [#78], okay, let me keep this simple for you. The price for questioning is not having an alternative explanation. If you were to question my assertions regarding quantization error not coming from an actual stochastic process, I would explain in more detail, giving references to scientific and engineering literature if necessary, and I wouldn’t demand that you first provide an alternative explanation.

    Only evolanders require questioners to have an alternative ready before addressing skeptical questions. The restaurant critic analogy is perfect, not irrelevant.

    Regarding [#79]‘s response to [#75], genes are transferred to descendants. Genetic osmosis among unrelated individuals is not a plausible “other mechanism” as far as I know.

    Hence, dmso74 still has a good point that you did not address, viz., that under the assumption that ERV integration cites are random, *and* the additional and implicit assumption that independent insertions are never able occur at the same location, then instances of identical insertion points in two different creatures *must* mean that the two creatures descended from the same *individual*.

    This second and implicit assumption may or may not have been an intended target of dmso74, but it is a key issue.

    Specifically, under this implicit assumption, it is not even possible to differentiate two unrelated individuals sharing an ERV at the same location due to coincidence from the hypothesized common descent conclusion. The common descent conclusion is a logical consequence of the implicit assumption, which is why I am asking for scientific justification for the assumption.

    [#82], your MRCA point is valid, but it in no way invalidates dmso74′s point. If an ERV is present in a population, every individual that has the ERV at the specific integration point, under the assumptions made above, must have descended from the original synthesizer (OS) of the RV. If the ERV is universal, than every individual in the population is descended from the OS. Even if that OS is not the MRCA, it is a CA. The OS can be a MRCA, or just a CA. But unless ERVs are spread through osmosis and not reproductively, dmso74′s skepticism, which seems to be against common descent in general, has not been addressed.

  86. #86 William Wallace
    April 13, 2009

    As a prophylactic against evolander debate tactics from getting out of control, and evolanders gloating over a typographical error: I meant site not cite.

  87. #87 Shrunk
    April 13, 2009

    @cprs (#80)

    Abbie’s already written a post on CERV 1 (also known as PtERV 1). It’s a fascinating story, if not a bit creepy (Zombie viruses! Brrr!), but the simple answer is our ancestors seemed to be immune to this virus, while chimps weren’t. Nothing to brag about, though; the tradeoff is that this has left us more susceptible to HIV.

    http://endogenousretrovirus.blogspot.com/2007/06/frankenstein-viruses.html

    With CERV 2, there may be someone here who know mores about this than I do, but it seems this could be simply explained by the ERV not being fixed in humans. This kind of touches on dmso74′s questions. Not every gene our common ancestor possessed is going to be fixed in a population (if so, we would be clones) and some of those genes could simply be lost entirely over time. It’s worth noting that, in the paragraph just before the section you quoted, the authors mention that CERV 2 was transpositionally active in chimps between 1.3 and 6 MYA, i.e. after the likely date of divergence between the human and chimp lines.

  88. #88 Albatrossity
    April 13, 2009

    WW

    I hate to break the news to you, but in science you really should have an alternative explanation in mind when you repeatedly question an explanation put forward by somebody else.

    As long as you understand that you are not behaving scientifically, I guess you can get away with your whines and diversions. But if you really want others to take you seriously and answer your questions, maybe you should pay attention to how science works. It’s not just a habit of your mythical evolanders; mathematicians, astronomers and other heroes of yours are respected precisely because they not only had questions, they had answers.

  89. #89 Shrunk
    April 13, 2009

    WW: your MRCA point is valid, but it in no way invalidates dmso74′s point. If an ERV is present in a population, every individual that has the ERV at the specific integration point, under the assumptions made above, must have descended from the original synthesizer (OS) of the RV. If the ERV is universal, than every individual in the population is descended from the OS. Even if that OS is not the MRCA, it is a CA. The OS can be a MRCA, or just a CA. But unless ERVs are spread through osmosis and not reproductively, dmso74′s skepticism, which seems to be against common descent in general, has not been addressed.

    You’ve got it partly right. True the OS is likely a CA. But what’s so special about that? It comes back to what I said to dmso74: This is basic population genetics. Even if you believe the human race just emerged from God’s rectum fully formed, the fact remains you only have to go back a few thousand years and the world is full of common ancestors. After a certain point (“after” in the reverse chronological sence of going backwards in time), if a person is an ancestor to anyone living today, then he/she is ancestor to everyone living today. It may be hard to get your head around that fact, but it’s simple mathematics and, as I said, would hold even if creationism were true.

    It’s already been said, but a look at the Wikipedia article on Most Recent Common Ancestor will make this clearer.

  90. #90 dmso74
    April 13, 2009

    Shrunk,

    Thank you, the wikipedia article did help (much more than “learn some population genetics” snark). i might have to do the math myself to make sure i really get it, but the family tree analogy helps.

    fyi i am an academic biologist (obviously lacking in pop gen training), and nowhere close to a creationist.

  91. #91 William Wallace
    April 13, 2009

    Albatrossity, if what you say is true, scientific journals would only publish letters critical of papers if the letters also posited alternative explanations. If scientist A wrote a letter merely pointing out how an assumption, made in a paper written by scientist B, taken to its logical conclusion would produce a paradox, for example, under your understanding of “science” said letter would not be published as it puts forth no alternative explanation.

    Forgive me if I don’t believe you. Indeed, it seems to me that journal editors would favor response letters that analyzed the paper in question, and did not try to sneak in alternative ideas via the letter route, as opposed to vetting those alternatives through the more formally peer reviewed paper path.

    WW

    P.S. This is a blog, not a journal.

  92. #92 Shrunk
    April 13, 2009

    fyi i am an academic biologist (obviously lacking in pop gen training), and nowhere close to a creationist.

    I realize that. Unfortunately, I accidentally posted an earlier edit of my response that inadvertently implied that you might be one. My apologies.

  93. #93 Shrunk
    April 13, 2009

    BTW, folks really should pop over to the YouTube link WW graciously provided us. It’s quite amusing watching him eat crow and claim victory at the same time.

    http://www.youtube.com/watch?v=viszQiV-0Us

  94. #94 Pete
    April 13, 2009

    I wonder if anyone has done any statistics determining the “mean time to insult” for the creation vs evolution debate, namely the average amount of responses before the entire conversation is reduced to exchanging insults. I’m guessing it is a low number.

    Then again it is probably not a creation/evolution thing, but an internet thing. So much easier to start insulting when your not face to face.

    Anyway, to try to stay on topic, I have a question for Mr. Wallace. I’m not sure what standard of evidence you would accept for randomness, but lets just hypothesize that Abbie does show the insertion is random, or indeed, perhaps not random, but that the virus will target one of 3000 or so entry points. Given that ERVs fall into a nested hierarchy pattern, creating a phylogenetic (sp?) tree that cross corroborates with the same phylogenetic tree created by alternate molecular evidence (such as mutations in a pseduogene), would this be evidence for you of common descent? What other explanation would you suggest can acount for this data? And finally ,obviously you are not convinced concerning common descent, but could you hypothesize what evidence you would be convinced by, if common descent were true, what would YOU expect to find in the pattern of genetics across multiple species?

    Thanks,
    Pete

  95. #95 cprs
    April 13, 2009

    Thanks for comments Shrunk – will pursue.

  96. #96 William Wallace
    April 13, 2009

    I hope [#91] isn’t ignored….

    Shrunk, Dr. Jackson has acknowledged an error, first pointed out by a Monday morning quarterback, in assigning the Morganucodon to the permian layer when it is in fact found in the triassic layers (the layers immediately above the permian layer). It is one point, found by a Monday morning quarterback, readily acknowledged by Dr. Jackson in an email to me.

    I don’t hold it too much against Dr. J., since he fist said the Permian layer was from 250 million years ago in evolutionary scale, but that the Morganucodon was from 210 million year ago. Both are about correct. But you may hold it against him if you wish. I like the video, too bad it is sullied by Dr. Jackson’s error in assigning the Morganucodon to the permian layers instead of the triassic.

  97. #97 Pete
    April 13, 2009

    I posted my question before I had read all the comments. Mr. Wallace, you seem to suggest in your sixth paragraph of post 85 that if the sites are random, common descent is a logical conclusion. Is this correct?

    Shrunk’s comments about population genetics are true, though I have a hard time wrapping my head around the idea as well. Dawkins explains it well in “The Ancestor’s Tale”, about how given enough time, if you are the ancestor of anyone you will be the ancestor of everyone. And yet, most likely ALL of your particular genetic material will have been weeded out. Indeed long term you might be the ancestor to everyone and yet contribute nothing to the genetic pool.

  98. #98 William Wallace
    April 13, 2009

    Pete,

    I don’t know if insertion points are random, or if there is a latent pattern, or if an ERV can be coincidentally integrated to the exact same spot in two different individuals.

    Good evidence in the context Smith’s post would be a paper analyzing insertion points and concluding that they are random.

    That would be a start. It is something I requested back in [#5], but still have not had provided by any of the evolanders here.

    Others may proceed as though they are random, and follow that path as though the assumption were true, and that is not in and of itself invalid. Such lines of inquiry are common in science.

    But asserting that they are random as part of an argument supposedly proving common descent, without some sort of scientific justification, is not science. It is sophistry.

    You start out with assumptions of random insertion points, and indeed, an implied assumption that that rules out the possibility of identical but coincidental integration points in different individuals, and then say, “see, they weren’t coincidental.”

    Regarding your other point, I am not convinced of macro evolution. I would be convinced if a repeatable experiment were devised to create a new species. Say, for example, a scientist were to perform a founder/flush cycle on some creature, creating a new population, splitting it, allow the two populations to go through a number of generations in a new environment with new selective pressures, where the number of generations is either fixed, or specified as a range, and after the successive generations are allowed to breed, one or the other population develops a new tangible physical trait, say antlers, or an additional nose, or a blowhole, that clearly marks it as a new creature.

    This sort of thing has been attempted, e.g., Dodd’s 1989 fruit fly paper describing her observations, but the best they could do as far as I understand it was generate more fruit flies. Sure, population A could eat Starch based food, while population B could eat maltose based food, and ~80% of the females surveyed from population A preferred to mate with males from population A, while only ~20% of the females surveyed from population A would bother to mate with males from population B. But they were all still fruit flies capable of and even willing to breed with one another.

  99. #99 pete
    April 13, 2009

    That would be a good experiment, though I would quibble on one point, if you started with a fish any new trait would in reality be a modification of something the fish already has, so antlers are probably out of the question until many many many different modified something else’s have taken place. Alas, sadly the experiment would probably not produce clear morphological differences for tens of millions of years or so. And once football season starts I would lose interest so we only have four months or so.

    But back to my point, you didn’t quite answer the question. Assuming for a second that the sites are random, AND (I suppose) assuming there is a nested hierarchy pattern (since you seem to suggest there might not be), assuming these things to be true would this be evidence of common descent?

  100. #100 Albatrossity
    April 13, 2009

    WW

    If you ever actually read letters in scientific journals you would know that scientists who pen such letters usually DO offer alternative explanations for hypotheses or conclusions that they disagree with.

    There are certainly letters expressing opinions about such things as conferences or nomenclature, but a letter written to disagree with a published primary paper almost always includes not only the factual disagreement, but also an alternative explanation for the observed facts.

    You really need to get out more into the biological science literature. It would save you from this sort of embarrassment.

The site is currently under maintenance and will be back shortly. New comments have been disabled during this time, please check back soon.