I think you’re clearly misrepresenting what Ann said. I assume that the misrepresentation was sincere, but it suggests a lack of charity or careful reading. She said that it did not look like a human design, because too sophisticated - not that it did not look designed. I think that this repeated claim of yours should be retracted.
One reason that I have little time for a lot of work by theistic evolutionists is that for a lot of people taking that view it seems that nothing is allowed to be taken as surprising on a naturalistic evolutionary account. This ignores biological reality in my view, which is full of surprises, both unexpected elegance and unexpected messiness - but, distinguishing between the two is often very difficult, so high levels of intellectual humility are warranted. e.g. is antisense transcription and/or translation (say, in bacteria) mostly/largely noise, or largely functional? We still don’t actually know.
I had hoped that Peaceful Science would foster serious discussions on the science, and to some extent it has (more than most other fora) but often discussions get derailed with petty sniping, which is a pity, and rather disappointing.
You may know that I work on overlapping genes, and am happy to say that I am one of very few experts on the topic. The truth is, as I see it, that they are highly surprising genetic constructs. In fact, it is very hard to tell how many there actually are, for various technical reasons which we are in the process of writing up.
There are a lot of unanswered questions in this area. I am working hard to answer some of them, with my students and other group members. The topic is interesting precisely because it is something of an enigma. I don’t think brushing over that surprise factor (which I take to be a key motivator for science) is helpful, any more than jumping to use them in an anti-evolutionary argument is (not that Ann actually explicitly used them to argue against evolution per se, as I recall).
In any case, while these genes do have an evolutionary history (which we are working on tracing) I think the genetic code’s ability to facilitate overlapping genes is highly remarkable - we have some papers coming out on this topic in the next few months.
I’m not sure if that is directed to this exchange or not.
For the record, there are a lot of surprising and non intuitive things in biology. We don’t understand many of them. I agree with that!
Overprinting was initially surprising to me, but now less so. It doesn’t make sense from an ID model of proteins though, and is good evidence against it.
This is blatantly a mischaracterization. I struggle to see how anyone could think otherwise.
She is saying that it doesn’t look like the way a human would design it, because we don’t design messages inside messages. That is too sophisticated for us. Therefore it requires a design genius.
That is her reasoning. I am not claiming to agree with her (as it seems that overlapping genes can crop up fairly easily - but I’m unsure!) but I understand why someone would make the claim that she does, and it needs to be fairly represented.
That is what I said. When it looks like human design, they take it as evidence of design. When it does not look like human design they still conclude design!
I’m not at all clear how I “blatantly misrepresented” her. In fact, I’m saying the same thing you just said.
You claimed to not be able to understand her reasoning.
It’s not complicated. She claims that the structure observed is highly sophisticated - NOT that it is a bad design or does not look designed, but that it is not the normal way that humans design things. It is the level of perceived sophistication that makes her say that it looks like the work of a design genius.
You can of course disagree with her regarding the sophistication, but it is not reasonable to say that the structure of her inference makes no sense. Ask anyone on the street whether designing a message inside a message is harder than designing a message by itself and they’ll say yes, of course.
I think in reality it’s not as hard as it looks, in part because of the remarkable structure of the genetic code. But, I’ll wait until the papers I’m working on on that topic are finished before I explore that further.
Perhaps reread her article and the quotes I gave? She does explain it does not look like a human design.
I understand the argument but dont understand how anyone would think it is valid or consistent with other arguements she has made.
Do you understand how this is a valid argument? You keep hedging, so I’d guess not.
Perhaps you can disagree with me but I do not think the structure of her inference makes sense. It looks like a special pleading to me. Maybe I’m wrong, but that is what it looks like to me. It’s not a snipe at her to explain it.
I’m not sure what you are getting at here. First, I agree that it doesn’t look like a human design, at least not as convieved by ID.
Second, is almost trivially easy to design a message within a message. The trick is picking an encoding with a lot of overlapping redundancy, such as an interleaved encoding.
Third the fact that overprinting is so common is evidence that there is not terribly high specification at the protein sequence level. Is that surprising? Well, anyone who agrees with Axe should be stunned, as this is not a problem resolvable with “genius.” For many of us that already knew that amino acid sequences are not well specified to function, it is not really that surprising.
If you want to understand why I find your critique of Ann problematic I think you should ask someone else who is more sympathetic to ID than you are to give you feedback on the way you have assessed this situation. I cannot explain it any more clearly.
I think that Ann was too strong in her claim that it doesn’t look like a human code, as humans have in fact designed such codes, for encryption of information, as I assume you also know. Her key claim though is that the situation of overlapping coding appears highly sophisticated, more so than most human designs. Your claim that this undermines her inference to design is not plausible. I think you have uncharitably interpreted two of her claims in such a way as to find a conflict, where no conflict is necessary at all.
We don’t know how common overprinting is, as I have said.
We also don’t know the proportion of protein sequence space which is functional, or folds, or whatever criterion someone may want. We just have a few estimates, some of which are disputed or disputable, using extremely different methods and testing different things.
These are difficult questions, and I am very familiar with the literature on them, and have heard many talks by leading experts who believe that functional proteins are very rare in sequence space. Proteins early in their evolution may well be different, but we are just barely starting to explore this.
I’m not impressed with the idea that people who understood it could have known all along that overprinting was easy. Firstly, because there are good biophysical reasons to doubt that folded proteins are common in sequence space (I suspect that overlapping proteins don’t adopt a highly specified fold, unlike e.g. advanced enzymes - but this is one of the many things we don’t have any real data on). And secondly because there is not a huge amount of excess redundancy in the standard genetic code - the redundancy that is there has been shown to be used for error correction and for optimising codon usage, and the code has further properties which have been shown to be optimised. I assume that you’re not too familiar with the literature on this. Allowing overlapping genes as well as these other things is quite remarkable, and does not strike me as at all easy.
@Zachary_Ardern you are free to disagree with me, and I appreciate the constructive resistance.
This exchange has clarified that we are looking at a disagreement between us, which is entirely okay. This not however me being unfair to Ann or ID. I just disagree, and I can explain why I disagree.
For that reason, can we please back off the claims of me being unfair? It would be more fun to sort out our points of agreement and disagreement. I might learn something that way!
It seems more like a reflex explanation to something she doesn’t understand. From Axe’s theory of protein space this is impossible. No matter how much genius is involved, 1+1 does not equal 3. Appealing to sophistication is a fallacious appeal.
Thank you for engaging, I appreciate it. I do still honestly think that you were being unfair, as I am quite sure there is nothing wrong with the structure of her inference that within a design framework a more sophisticated design implies a more intelligent designer, and I feel that you did not accurately present the argument that she was offering.
(NB: I keep noting that I don’t agree with the actual argument as applied to this case, or “hedging” as you said, because I disagree with the implicit empirical claim that these sequences cannot evolve and are directly designed. Also you never know who will read these comments one day, and I want to make it quite clear to them that I am not a critic of evolution)
The main problems are subjectivity and lack of falsification. @agauger is simply giving her subjectivity opinion of what looks designed or sophisticated which means it doesn’t have any empirical or objective data that science can grab onto. Next, we would suspect that anything Ann sees in biology will looked designed to her, for whatever reason. That tends to be the case when you start with a conclusion. On top of being a subjective opinion, it also gives the vibe of being a very biased one.
The first assumption we would need to address is whether specific folding is necessary for protein function.
Perhaps I am missing something, but it doesn’t seem surprising that there can be alternate reading frames, or why this would be any more sophisticated than a lack of alternate reading frames. Mutations that promote transcription or translation at alternate sites will pass through the same selection filter that all mutations go through. In fact, I think it would be more sophisticated to have a transcription/translation system that was so specific that alternate reading frames could not emerge through random mutations.
After three years of working in depth on this topic, I am happy to affirm that overlapping coding (if it happens much - it may be extremely rare) is a highly sophisticated way to encode genes. This has got nothing to do with something that I don’t understand, but what I do.
I would be happy to bet that Ann knows at least as much as you do about overlapping coding, as well.
Ann knows that Axe’s figures don’t apply precisely to all proteins. The actual numbers across proteins are unknown, but many experts believe that folded proteins with a given function are very rare in sequence space (less than 1 in 10^-10 is a figure I heard from one indubitable German expert on protein evolution in a talk last year), so the “function is rare” perspective is not dumb at all.
The existence of overlapping (i.e. alternative) open reading frames per se is not at all surprising. Being able to both encode and evolve useful proteins in many of them, while retaining function in the originally-encoded-frame does appear surprising. Relevant factors seem to include reasonable ORF length, suitable clustering of amino acid properties, and getting a balance between conservative and explorative mutation effects following mutation in the originally-encoding-frame. The details of these things will come out in future papers from my group.
While the number (1 in 10^-10) sounds daunting, it isn’t so small as to preclude the origination of new proteins, or even the occurrence of functional overlapping genes. Probably more important for a discussion of ID (as this forum often is), this number is quite inconsistent with the “function is rare” mantra.
Zachary, it looks to me (from a quick reading) that your focus is on bacterial genomes. I was wondering what your thoughts are of the uses of overlapping open reading frames by RNA viruses. Are there parallels to be found with bacteria? Do you think some of the considerations we are discussing might help understand the specific biochemical functions of some of these proteins?
How much she knows has nothing to do with how much is understood of something.
I don’t understand this at all. How did you measure “sophistication”? What objective metrics of sophistication did you use? What is the literature out there on the sophistication of overprinting?
I’m not being difficult, i just have no idea what it means to say this.
If we are to believe ENCODE’s definition of “function”, then it seems to happen very frequently in humans, and in cancer. I can give you references if you like. It is not extremely rare, just usually very low amounts.
It happens very frequently at high amounts in viruses. Also, we know why it happens frequently in viruses.
You are reading her mind here, as she did not say that in the article. Noting the conflict with Axe’s numbers, and explaining that Axe’s numbers didn’t apply here would have been a better more coherent response. I would have given her credit for noting the limitations of Axe’s numbers and the conflict that overprinting presents to his numbers.
She did not do this. If she knows this to be true, she should say it. Can you produce any references where she has clarified this in relation to overprinting? If not, why won’t she say it if you read her mind correctly?
To whom and under what model? It certainly is surprising if you think that DNA is precisely a language or a computer code (as we often hear from ID). It is not that surprising to those that understood the limitations of the analogy.
I’ll look forward to those papers! Some of this has been published already right? For example, antisense frames of coding DNA have longer ORFs just as a property of the genetic code.
Well, I suppose we disagree. I think the structure of her argument is clearly fallacious. She doesn’t understand overprinting under her model, so she calls it “genius,” which is impenetrable. Under Axe’s model of protein space (to which she subscribes) this should be impossible, not “sophisticated.” No amount of genius makes 1+1=3. Face with that contradiction, an appeal to genius is a special pleading, not a coherent argument.
And to be clear, I did very accurately present the argument. It matches your summary! I accurately summarized her argument, and also explained why it I though ti was fallacious. You think the structure of the argument is valid (why?), but still disagree (noted). We both agree she is wrong then, but for different reasons. I’m not sure what you are objecting to.
I am very familiar with this @Zachary_Ardern. I’ve done work in this area and advised several students on this.
The key observations that untangles the mechanistic knot is the difference between viruses and eukaryotes. Do you know why overprinting is more common in viruses than eukaryotes? Do you know which classes of virus have increased overprinting and why?
The reason why is, primarily, about fuzziness in definitions. We do actually know how common overprinting it is viruses and in humans, but there is a continuous range of definitions.
As we can see, there is a large amount of substantive disagreement. And I’m not as uninformed as you seem to think I am. Perhaps I am wrong, but I’m not ignorant. Perhaps I am wrong, but I am engaging you and Ann.
So let’s deal with the substantive issues. Maybe you can teach me something! Maybe we can help you with your project!
Let’s start wiht this: what are some of the more intersting things you’ve found in your study so far?
Once again, the patterns of differences in overprinting in all these contexts gives a lot of information about how overprinting arises and how it is maintained or lost.