Stairways to Understanding (tentative books project)

I’ll try to dig something up. Meanwhile there’s this: Opsin evolution: orgins of opsins - genomewiki

  • Opsins are definitely not the ‘original’ GPCR because these were already widely deployed at much earlier divergence nodes – yeast, protozoa, choanflagellates, trichoplax have GPCR but lack opsins. Nor are opsins the prototype for the ‘rhodopsin class’ R of the GRAFS classification of GPCR which again was established far earlier. Indeed, even the Ralpha subgroup with of rhodopsin class GPCR was well-established prior to the first metazoan opsin.

  • Opsins are thus latecomers, not pioneers, to a rapidly expanding paralogous gene clade within already full-featured GPCR. Judging by their closest extant blastp relatives among tens of thousands of GPCR at GenBank, opsins specifically arose as a gene duplication within the peptide receptor subgroup PEP. Indeed, certain of these proteins list opsins among their top ten best back-blast matches (ie have better matches than to almost all non-opsin GPCR). Note here that blast scores can be misleading because the ‘floor’ of percent identity is about 25% just due to universal conserved residues plus accidental matches.

It appears I must eat my own words, as I can’t seem to find any specific references that suggest opsins derive specifically from GPC receptors classified as olfactory receptors(in fact I found a reference that explicitly stated the diametrically opposite)*. And I can’t remember where exactly I got this idea from, so I suppose it could just be a false memory and I’ve got something mixed up.

Rather the evidence appears to suggest they’re more closely related to hormone and peptide binding GPC receptors that could just as well have functioned in cell-cell signaling within a multicellular organism. This is also discussed further down in the link I gave.

*They write: The evolution of opsin genes is in sharp contrast to that of OR genes. Opsin genes also belong to the rhodopsin-like GPCR superfamily, and thus they are distant relatives of OR genes. Homologous opsin genes can be identified in vertebrates, insects and other invertebrates, however, indicating that the origin of the opsin gene family is much more ancient than that of the OR gene family[58].


Could you explain what that colorful figure means?

It means just because we can pump sequences into computer and generate a nested hierarchy, it doesn’t mean the nested hierarchy was the product of random mutation and natural selection.

Having an almost correct Zinc Finger array (there are about 13 Zinc Finger in my colorful figure) is like having an almost correct computer password. There aren’t really good gradual evolutionary routes to a coherent zinc finger array that targets specific locations in the genome anymore than there are gradual routes to figuring out a computer password.

Each of the Zinc Fingers of about 30-32 residues correspond approximately to a target of 3 DNA bases, but it’s not as modular a design as synthetic chemists would like since a given Zinc Finger modifies the chemical behavior of the adjacent Zinc Fingers. This illustrates the password problem with a mere 3 Zinc Fingers:

Proteins like the KRAB Zinc Finger (KRAB KZNF in the diagram) attach to segments of DNA (often ERVs or other elements) to help recruit molecular machines:

The binding of KRAB-ZFP (KRAB Zinc Fingers) to ERVs or other elements means the Zinc Finger array must be matched to recognize a DNA target. The problem for evolutionary theory was that to have a nested hierarchy across species of KRAB-Zinc Fingers implied a corresponding nested hierarchy of targets!

I found that absurd, but nevertheless, evolutionary biologists declared this as some sort of evidence of co-evolution, which struck me as nothing more than an epicycle.

Anyway here is the related paper that talks about the co-evolution KRAB-ZFPs with evolution of a certain kind of transposon. It describes the problem I was mentioning, but I don’t agree with the “solution” of co-evolution. I’d sooner believe in miracles.

Sorry for the delay in responding to your good question.

It isn’t my preferred hypothesis, but it would be my 2nd choice, that is if we’re talking Progressive/Old Earth Creationism which was (is?) Stephen Meyer’s view according to my interview with him in around 2010. Here is a link to my interview with him:

The 3rd choice would be Behe’s common descent with ID along the way, but how that is not different from miracles is not something Dr. Behe delves into.

But specifically to your question, the Faint Young Sun Paradox was the beginning of my doubts to the age of the fossil record. If the fossil record is young, then there was no time for progressive creationism anyway.

For the Faint Young Sun Paradox to be resolved it requires miracles of fine-tuned global warming. I suppose that’s a lesser miracle than special creation of all species, but still, it was my beginning of doubts for the age of the fossil record – doubts which obviously persist to this day, now also reinforced with the strong evidence the human genome is deteriorating and thus could not have evolved naturally. As Kondrashov said, “Why aren’t we dead 100 times over”?

2 posts were split to a new topic: Criticisms of Stairways to Understanding

No, please try again. Where is the figure from? What do the arrows represent? What sequences are shown at right? What data were used to build the tree? What algorithm?How robust is the tree? If you did that yourself, I’m suspecting that you aren’t quite clear on what you did and what it means. But at the moment it’s certainly unintelligible to other people.

That makes a sort of sense, but if the fossil record is young, there’s no reason for any correspondence between the record and phylogeny. You’re allowing one little anomaly to overrule all the rest of the data. Bad practice.

There is no such evidence, but if there were, shouldn’t it apply to all species, and shouldn’t the faster-reproducing ones be all extinct by now?


The protein in the one of the diagrams is described here.

Note the identifier “P52737” is also in my diagram. It gives the same name as in the diagram. The website provides the same FASTA sequence as in the diagram.

The coordinates of the zinc fingers are also there at that website:

|Zinc fingeri|140 – 162|C2H2-type 1; degeneratePROSITE-ProRule annotationAdd BLAST||23|
|Zinc fingeri|168 – 190|C2H2-type 2PROSITE-ProRule annotationAdd BLAST||23|
|Zinc fingeri|196 – 218|C2H2-type 3PROSITE-ProRule annotationAdd BLAST||23|
|Zinc fingeri|224 – 246|C2H2-type 4PROSITE-ProRule annotationAdd BLAST||23|
|Zinc fingeri|252 – 274|C2H2-type 5PROSITE-ProRule annotationAdd BLAST||23|
|Zinc fingeri|280 – 302|C2H2-type 6PROSITE-ProRule annotationAdd BLAST||23|
|Zinc fingeri|308 – 330|C2H2-type 7PROSITE-ProRule annotationAdd BLAST||23|
|Zinc fingeri|336 – 358|C2H2-type 8PROSITE-ProRule annotationAdd BLAST||23|
|Zinc fingeri|364 – 386|C2H2-type 9PROSITE-ProRule annotationAdd BLAST||23|
|Zinc fingeri|392 – 414|C2H2-type 10PROSITE-ProRule annotationAdd BLAST||23|
|Zinc fingeri|420 – 442|C2H2-type 11PROSITE-ProRule annotationAdd BLAST||23|
|Zinc fingeri|448 – 470|C2H2-type 12PROSITE-ProRule annotationAdd BLAST||23|
|Zinc fingeri|476 – 498|C2H2-type 13PROSITE-ProRule annotationAdd BLAST||23|
|Zinc fingeri|504 – 526|C2H2-type 14PROSITE-ProRule annotationAdd BLAST||23|

One could add the amino acids between zinc fingers to make the rows in my diagram. The colored columns indicated conserved residues or conserved classes of residues (like aromatic residues).

The arrows point to a hierarchy generated by MEGA by taking the individual zinc finger rows. Note, I even put the FASTA coordinates in the tree, identifying the protein by name and the coordinates of the particular zinc finger and the ZF number as designated by UNIPROT.

Whether one uses rooted or unrooted neighbor joining, maximum likelihood, bootstrapped or not, …the point was all of these would be pretty much fictions because of the problem of creating a functional binding as the zinc finger array evolved.

Whatever hierarchical structure derived by supposed phylogenetic methods is only an imagination in the mind of the phylogeneticist unless he can justify the functional of the protein in each of the evolutionary phases it goes through.

The point is, making hierarchies isn’t proof of evolution, it’s proof things can be arranged hierarchically! The ZN136 protein is an example where it would be absurd to think this actually evolved because the protein would be functionally broken in almost every step of evolution.

The problem of hierarchy of domains in a single protein can be extended to paralogs in a single organism and then to orthologs across organisms. Just because one can make a hierarchical tree doesn’t mean that it can reasonably be evolved. Zinc finger proteins such as in the KRAB family are counter examples to that belief.

This is a model of zinc finger domains binding to 3 DNA bases per zinc finger:

In light of this diagram it is hopefully apparent that random changes to zinc fingers wouldn’t be good.

One only needs to scour medical literature to see the consequences of mutations to the zinc finger array.

If that’s the case, why should it be expected randomly emerged zinc finger arrays would be selected for at all at any stage of evolution of the zinc finger protein?

To clarify for the reader’s benefit, here is a diagram that includes the FASTA file of the Human ZN136 protein. It shows the rows in the FASTA file that implement a valid classical C2H2 Zinc Finger motif. I show the conceptual fold of the Zinc Finger domain, as well as a more realistic 3D fold of the domain:

Turning the FASTA file on it side, I show conceptually how the fold of the multliple zinc fingers correspond to the sequences:

conceptually this is how a multizinc finger protein might bind to DNA:

The TFIIIA protein, btw, binds to 55 bases. It is a little different than other zinc finger proteins, and doesn’t exactly conform to the 3-bases-per-zinc finger rule.

But hierarchical clustering is a great way of understanding relatedness. For example, this technique of tracking mutations to show relationships is common in immunology: showing the relatedness of clones of B cells or T cells that have mutated and selected as they develop higher affinities for their target antigen. (In a sense, these immune cells are “evolving” under selective pressures in their environment, and developing higher affinity for their target antigens as immune cells with detrimental or unproductive mutations die off)

See figure 5 in this paper, as an example:
Salmonella Infection Drives Promiscuous B Cell Activation Followed by Extrafollicular Affinity Maturation

I presume the hierarchical clustering technique is also how scientists are currently tracking the transmission of the covid-19 virus, allowing them to determine if the virus came to NYC from Europe or directly from China, for example.

So I do believe that comparing mutations across phylogenic trees is a powerful evidence for common descent.


No, sorry. You can make a tree out of any data at all. The question is whether the tree has good support, i.e. whether the data really do form a nested hierarchy or you have just imposed one. The shorter the sequence and the worse your model, the less likely you are to find good support even if the hierarchy is there.

Then again, the literature is fully of very strongly supported trees showing real nested hierarchies, not just those imposed by the analysis. Can you agree that common descent is an obvious explanation for those hierarchies? Can you come up with another plausible explanation?

Have you ever consulted the literature on the evolution of ZN136, incidentally?



Now look at ZNF136 (aka ZN136) between humans and pigs.

This is a BLASTP alignment

The human zinc finger array looks different than the pig zinc finger array.

If we said those are just neutral mutational differences, well maybe OK, BUT recall the zinc fingers have binding partners, namely DNA, and DNA that does NOT code for the zinc finger protein!

Recall the binding partners of a KRAB zinc finger array is DNA. Conceptually the ZNF136 in humans binds to targets in human genome (in red) and the ZNF136 in pigs bind to targets in the pig genome (in blue). The red and blue DNA is not likely to be the same because the from the BLAST sequence above, the zinc fingers which target these segments of DNA are spelled differently.

This is what is meant in the literature that KRAB zinc finger proteins co-evolved with their targets.

Therefore, whatever hierarchical construction that is made for ZNF136 (NJ, maximum likelihood, maximum parsimony, rooted, unrooted, boostrapped or not, etc.) across mammalian species cannot be interpreted as a real common descent phylogeny unless one invokes statistical miracles of co-evolution. Reductio ad absurbdum.

Why would we say that in the first place?

Yes, obviously.

I’m sorry, but your conclusion doesn’t follow from your premises. You don’t deserve a “therefore”. You have not shown that there is any statistical miracle needed. There’s some gross misconception of evolution lurking there, but I don’t have enough information to tell me just what. But coevolution involves selection, not just neutral evolution.


Yes, in a classic example of an evolutionary arms-race. One molecule A that is the binding target for another molecule B, mutates, then to adapt B has to mutate to keep binding it with the same affinity and strength as before.

That’s exactly the same basis of the evolutionary arms-race between the adaptive immune system evolving to adapt to each season’s (for example) novel influenza virus mutations.

That perfectly well explains why the Zinc finger protein sequence, and it’s DNA binding target in human and pig DNA are not identical.

What in the world are you talking about? This is the Chewbacca defense of creationism.

You put a therefore as the first word in that sentence, but noting it states is even weakly implied by any of the preceding discussion you did. The diametrically opposite is the case.

Look, you can’t just mindlessly point to some obscure factoid, and then invent a completely disconnected entailment.
Ladies and gentlemen, that man is wearing a coat. That man over there is wearing a coat. Why would he be wearing a coat? It’s simple, he wouldn’t! Therefore evolution is impossible. You must acquit!


This model does not apply to all zinc finger domains.

1 Like

I’m looking at the zink finger protein with it’s sequential, modular arrangement of individual zinc finger loops, and it is immediately obvious how this thing intrinsically lends itself to step-wise evolutionary change.

Segmental duplications, rearrangement of, and substitutions in these loops all obviously have the capacity to alter the DNA binding affinity and specificity of the molecule. It’s difficult to think of another polymer macromolecule that could more obviously fit a modular LEGO block-style analogy of incremental change better than this.


Let’s apply Sal’s argument to a closely related field: human ancestry.

An Approach to Human Ancestry That Is Logically Consistent with Sal's Protein Phylogenetics Argument

“Human ancestry is commonly supposed to have a hierarchical structure. Whatever hierarchical structure derived by supposed phylogenetic methods is only an imagination in the mind of the ancestral researcher unless he can justify the names and places of birth in each of the generations it goes through from Adam and Eve to the present day. If you cannot provide the key information about each of your ancestors who was alive at any point in the past 6000 years, your claim of descent from Adam and Eve is 100% in error!”

“The point is, making hierarchies isn’t proof of human ancestry, it’s proof things can be arranged hierarchically!”


Thank you for your comment. Are the B cell and T cell clones described here from or within the same organism?

That’s not my argument.

Humans can mate with each other and make other humans, therefore the hierarchy is meaningful if we are surveying only human DNA because the premise of common ancestry within the human lineage is reasonably well established.

In the other cases, existence of hierarchy alone uses as PROOF of common ancestry, and hence the reasoning is circular because it doesn’t consider the mechanistic feasibility of common ancestry. Of course one could still assert common ancestry by invoking miracles to make it possible for a Lemon Plant to have a common ancestor with a frog, but if one invokes miracles, how is that materially different from creationism. At least creationists admit miracles in their model, evolutionists don’t admit miracles even when it is obvious the theory needs miracles to rescue the theory.