Common Design vs. Common Descent

I promised John Harshman for several months that I would start a discussion about common design vs. common descent, and I’d like to keep my word to him as best as possible.

Strictly the speaking common design and common descent aren’t mutually exclusive, but if one invokes the possibility of recent special creation of all life, the two being mutually exclusive would be inevitable.

If one believes in a young fossil record (YFR) and thus likely believes life is young and therefore recently created, then one is a Young Life Creationist (YLC). YEC (young earth creationists) are automatically YLCs but there are a few YLCs who believe the Earth is old. So evidence in favor of YFR is evidence in favor of common design over common descent.

One can assume for the sake of argument the mainstream geological timelines of billions of years on planet Earth. If that is the case, special creation would have to happen likely in a progressive manner. I believe Stephen Meyer and many of the original ID proponents like Walter Bradley were progressive creationists.

Since I think there is promising evidence for YFR, I don’t think too much about common design vs. common descent. If the Earth is old, but the fossil record is young, as far as I’m concerned the nested hierarchical patterns of similarity are due to common design.

That said, for the sake of this discussion I will assume the fossil record is old. But even under that assumption, I don’t see how phylogenetics solves the problem of orphan features found distributed in the nested hierarchical patterns of similarity. I should point out, there is an important distinction between taxonomic nested hierarchies and phylogenetic nested hierarchies. The nested hierarchies I refer to are taxonomic, not phylogenetic. Phylogeneticsits insist the phylogenetic trees are good explanations for the taxonomic “trees”, but it doesn’t look that way to me at all. I find it revolting to think giraffes, apes, birds and turtles are under the Sarcopterygii clade (which looks more like a coelacanth).

Phylogeny is a nice superficial explanation for the pattern of taxonomic nested hierarchy in sets of proteins, DNA, whatever so long as a feature is actually shared among the creatures. That all breaks down however when we have orphan features that are not shared by sets of creatures.

The orphan features most evident to me are those associated with Eukaryotes. Phylogeny doesn’t do a good job of accounting for those. In fact, to assume common ancestry in that case, “poof” or some unknown mechanism is indicated. If the mechanism is unknown, then why claim universal common ancestry is a fact? Wouldn’t “we don’t know for sure, but we believe” be a more accurate statement of the state of affairs rather than saying “universal common ancestry is fact.”

So whenever orphan features sort of poof into existence, that suggests to me the patterns of nested hierarchy are explained better by common design. In fact there are lots of orphan features that define major groups of creatures. Off the top of my head, eukaryotes are divided into unicellular and multicellular creatures. There are vetebrates and a variety of invertebrates. Mammals have the orphan feature of mammary glands. The list could go on and on for orphan features and the groups they define. Now I use the phrase “orphan features” because I’m not comfortable using formal terms like autapomorphy or whatever. I actually don’t know what would be a good phrase.

So whenever I see an orphan feature that isn’t readily evolvable (like say a nervous system), I presume God did it, and therefore the similarities among creatures that have different orphan features is a the result of miraculous common design not ordinary common descent.

5,163 thoughts on “Common Design vs. Common Descent

  1. Mung:
    Salvador’s ramblings. Common Descent or Common Design?

    Random mutation and no selection whatsoever.

    😉

  2. There are what is known as “promiscuous domains” for proteins. Analogous to the complication that HGT creates for phylogeny, promiscuous protein domains create a problem for trying to classify protein based phylogeny rather than structure.

    But first here is a discussion of promiscuous protein domains:

    https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2722818/

    A substantial fraction of eukaryotic proteins contains multiple domains, some of which show a tendency to occur in diverse domain architectures and can be considered mobile (or ‘promiscuous’). These promiscuous domains are typically involved in protein–protein interactions and play crucial roles in interaction networks, particularly those contributing to signal transduction. They also play a major role in creating diversity of protein domain architecture in the proteome. It is now apparent that promiscuity is a volatile and relatively fast-changing feature in evolution, and that only a few domains retain their promiscuity status throughout evolution. Many such domains attained their promiscuity status independently in different lineages. Only recently, we have begun to understand the diversity of protein domain architectures and the role the promiscuous domains play in evolution of this diversity. However, many of the biological mechanisms of protein domain mobility remain shrouded in mystery. In this review, we discuss our present understanding of protein domain promiscuity, its evolution and its role in cellular function.

    There is a promiscuity score for these domains. The mechanism for them poofing into protein lineages is mysterious.

    But the essential point is that it makes sense to classify proteins based on structure and function, not phylogeny. With the case of muti-domain proteins coming from promiscuous ancestors (pun intended), it make sense to classify them on the present-day structure and function, not on who their multiple promiscuous ancestors were.

    The problem with promiscuous domains, like HGT is that one can’t just say a protein evolved via point mutations and indels without accounting for insertion of a promiscuous domain possibly from a gene far removed.

    Chapter 10 of Joe Felsenstein’s book added a lot of clarity to the various schools of thought and schools of phylogeny. He discusses Hennig’s work with morphology to infer phylogenies vs. the work of statistical inferences on gene trees. I can’t summarize Joe work and do it justice. Chapter 10 gives me reasons to reject using Hennig as a means of taxonomic classification.

    As can be seen by the work being done at the molecular level, classification (aka taxonomy) is purely structure based for protein domains, not really phylogeny based. It is a little superfluous to try to define a protein by its major common ancestor since some important properties may arise from contributions from promiscuous ancestors that combined to make a mutli-domain protein.

    Hence, I find it not very useful to invoke Hennig’s approach to taxonomy because it appears to fall apart in protein taxonomy. Basic similarity comparisons seem a superior way to do taxonomy and classification vs. using phylogenies (which are uncertain).

    Food for though regarding protein domains:

    Numerous sequence- and structure-based domain databases enable protein domain detection with very high accuracy, such as Pfam [3], SMART [4], CDD [5], INTERPRO [6], SCOP [7], ProDom [8], DALI [9] and CATH [10]. These databases either use sequence- or structure-based methods to identify regions in protein sequences that belong to specific domain families.
    ….
    While some domains are present only in specific combinations, others participate in diverse domain architectures. Domains of the latter types are called ‘promiscuous’ or mobile domains, and are very important in creating the observed diversity in protein domain architectures. They play a major role in signaling network in the cell by bringing together domains with different functionalities into one protein sequence, and thus promoting crosstalk in signaling.
    ….
    The number of unique domains in an organism is roughly proportional to its genome size. In unicellular eukaryotes, such as apicomplexans, diplomonads and protozoans, the unique number of domains is ∼1000, whereas in plants, fungi and animals, the numbers can be as high as ∼3000. The average size of domain is ∼100 amino acids [11]. The number of domains per gene (modularity) follows the power-law (see below) distribution [12], and it has been shown that tissue-specific genes have higher modularity [12, 13].

    The estimation of the frequency of multidomain proteins in the three superkingdoms of life (bacteria, archaea and eukaryotes) varies with the methodologies and database used [14–18], but the emerging consensus is that prokaryotes have fewer multidomain proteins than eukaryotes. The tendency of formation of multidomain proteins increases from archaea to bacteria to eukaryotes [1, 19]. Although within eukaryotes, particularly in animals, there is a distinct tendency towards formation of multidomain proteins (39% of metazoan proteins contain more than one Pfam domain, whereas the corresponding number for unicellular eukaryotes is smaller, 32% [20]), a large fraction of the proteins in all three super-kingdoms of life contain 0–1 domain [2, 18, 20, 21].

  3. From the paper earlier referenced:

    The propensity of protein domains to form multidomain architecture increases with organismal complexity. Though complexity is a contentious issue in evolution, here we define it as the number of cell types in an organism. The phenomenon that organisms with higher complexity tend to acquire more multidomain proteins is called ‘domain accretion’ [22], which could translate into increasing interaction amongst the domains. This may be one of the explanations of the apparent lack of correlation between the complexity and number of genes in a genome (G-value paradox): flies have fewer genes than nematodes; humans have fewer genes than rice [23]. Increasing modularity through domain accretion, at least in theory, can overcome the shortcoming posed by fewer genes in the genome. The biological mechanisms dictating domain accretion is not known. But, there is evidence that domains involved in the same functional pathway tend to come together in one protein sequence [24]. This phenomenon has been used to determine the functions of unknown domains in proteins, in what is called the ‘Rosetta Stone’ approach [24].

    It is worth mentioning this as a possible mechanism of promiscuous domains since some domains reside on exons:

    https://en.wikipedia.org/wiki/Exon_shuffling

    Exon shuffling is a molecular mechanism for the formation of new genes. It is a process through which two or more exons from different genes can be brought together ectopically, or the same exon can be duplicated, to create a new exon-intron structure.[1]

    The above uses the word ectopically:

    Ectopic recombination is an aberrant form of recombination in which crossing over occurs at non-homologous, rather than along homologous, loci. Such recombination often results in dramatic chromosomal rearrangement, which is generally deleterious.[1] Some research, however, has suggested that ectopic recombination can result in mutated chromosomes that benefit the organism.[2] Ectopic recombination can occur during both meiosis and mitosis, although it is more likely to occur during meiosis.[3] It occurs relatively frequently—in at least one yeast species (Saccharomyces cerevisiae) the frequency of ectopic recombination is roughly on par with that of allelic (or traditional) recombination.[4] If the alleles at two loci are heterozygous, then ectopic recombination is relatively likely to occur, whereas if the alleles are homozygous, they will almost certainly undergo allelic recombination.[4] Ectopic recombination does not require loci involved to be close to one another; it can occur between loci that are widely separated on a single chromosome, and has even been known to occur across chromosomes.[5] Neither does it require high levels of homology between sequences—the lower limit required for it to occur has been estimated at as low as 2.2 kb of homologous stretches of DNA nucleotides.[4]

  4. How do protein domains relate to the question of common design vs. common descent? If exon shuffling or other ectopic recombinations are inadequate explanations for promiscuous domains, especially domains unique to a species, it would seem Common Design via a miracle is a better explanation than common descent.

    The paper above alluded to the existence of novel protein domains in groups of organisms (like plants). Orphan genes are a challenge for evolution, but domains unique (dare we call them orphan domains) to a taxonomic group spread across numerous genes , are also a challenge. Too early to say how big the challenge is, but this is a compelling question to explore using bio-informatics and structural biology.

  5. stcordova: But the essential point is that it makes sense to classify proteins based on structure and function, not phylogeny.

    The two are not mutually exclusive. We humans are gifted, we can do both.

  6. The aspect of domain shuffling and exon shuffling obvious have some relation to Alternative Splicing a topic colewd is fascinated by.

    Protein domains are subtle and often the motif is really convincingly confirmed by actual structural analysis (like that done by X-ray crystallography) of the protein, not by mere sequence comparisons.

    But if the position and presence of protein domains are influenced by alternative splicing, this adds a level of potentially important complexity, and hence makes sense of one of the ways eukaryotes, especially multicellular eukaryotes, leverage exons.

    Unfortunately, the study of alternative splices and effects on different domain changes on the same protein are in it’s infancy. A representative paper (highly technical) on such issues is this one which focuses on a single protein:

    https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3481270/

    We conclude that TRPM3ΔICF [alternatively spliced] variants are regulatory channel subunits fine-tuning TRPM3 channel activity.

    This goes against Larry Moran claiming alternative splices are noise (at least as far as the proteome). Alternative splices in the transcriptome that don’t go to translation may have transcriptome roles. There is really too much we don’t know for Larry to be making pronouncements that this is all junk.

  7. Sal continues with the Gish Gallop. Sal, how does common design explain the nested hierarchy of life?

  8. John Harshman,

    Ignored
    Sal continues with the Gish Gallop. Sal, how does common design explain the nested hierarchy of life?

    Is there a clear definition of the nested hierarchy of life?

  9. John Harshman:

    Sal, how does common design explain the nested hierarchy of life?

    It explains it by saying, “God did it because common descent can’t.” You’re inability to rebut the conflicting data points is evidence common descent can’t do the job.

    Random mutation on genes doesn’t work so well for explaining spliceosomal introns. Universal common ancestry doesn’t explain the origin of spliceosomal introns very well.

    So what if I’m galloping around, it just highlights how many things you can’t explain with universal common ancestry.

    I notice you complain of a Gish Gallop when you are confronted with data that don’t agree with your models. I guess that’s the best you can do when you really don’t have an answer. Rather than admit you don’t have a credible mechanistic model, you make accusations of a Gish Gallop on me when the absence of explanations really falls on you. It’s a nice way of avoiding the fact you don’t have an answer. Just admit it, you’re guessing and believing your model just like I am. It’s a statement of faith without direct facts much like I have statement of faith without direct facts. You just won’t admit it you’re commitment to your model is faith-based as much as mine.

    Oh I failed to acknowledge earlier you corrected me on cycads not being ferns (they are fern-like). Thanks for the data point. Every now and then you do make a legitimate point. So I credit you for that.

    Nevertheless, the cycads show a molecular clock problem similar to the RNA virus clocking problem….

    That paper on RNA viruses being recent seems really interesting.

  10. colewd:

    Is there a clear definition of the nested hierarchy of life?

    Depends on who you ask. I would say structurally and morphologically that there is a nested hierarchy. Creationists were among the first to note this.

    Evolutionists have claimed the nested hierarchy of structures is proof of Universal Common Ancestry, but one could also argue the nested hierarchy of structures is evidence against macro evolution. How? Mammals come from other mammals, birds from birds. Birds don’t come from fish. We should expect fish give rise only to other fish. That is a structural argument against Macro Evolution. I also showed that Cox1 agrees with that claim.

    Nested hierarchies of structure leads to the chicken and egg paradox, if mammals give rise only to mammals, what gave rise to the first mammal?

    Here is a slanted account of the development nested hiearchies by creationists:
    https://evolution.berkeley.edu/evolibrary/article/history_05

    First the good part of nested hierarchies in terms of common design:

    Homo sapiens, Tyrannosaurus rex, Escherichia coli—our English conversation is littered with pairs of Latin names for animals, plants, and microbes. How did a dead language find this renewed life? It is the 250-year-old legacy of a Swedish naturalist’s quest to discover God’s handiwork in nature.

    Linnaeus
    Carolus Linnaeus (1707-1798) was far from the first thinker to try to classify life. Aristotle, for example, argued that each species had a unique form and could be classified by some of its key characteristics. In the process, he organized life in a ladder-like hierarchy, with plants on the bottom, animals in the middle, and humans on top (figure, right). Medieval European scholars were guided by both Aristotle and the Bible, and they believed that nature—including all of the species on Earth—reflected God’s benevolent organization of the world.

    Now the BS part of nested hierarchies in terms of Darwin:

    Biologists still use Linnaeus’ conventions today when they name a new species. But Darwin rendered the ideas behind those conventions obsolete. Darwin recognized that evolution could produce the hierarchy of similarities that so impressed Linnaeus, as old species gave rise to new species. Biologists still place pigs, porcupines, and people in Mammalia, but they do so because all the evidence—comparisons of fossils, anatomy, and genes— confirms that they descend from a common ancestor.

    You alluded earlier to the problem of convergence in bats and whales. The protein domain issue could make things very much worse for advocates of random mutation and common descent. We’ll see.

    But with respect to Alternative Splicing, upon connecting them to protein domains (as in lego building blocks), it now starts to make some sense how Alternative Splicing is important and why Humans and Chimps are different because of the different splices. Up until recently, I didn’t connect the dots.

    Now if the promiscuous domains that are unique to species groups are prevelant enough, this is a problem for advocates of common descent explaining the patterns of diversity in Gene Trees.

    I alluded to the problem of functional constraints on genes earlier:

    Proteome (post translational modifications: phospho proteome, acetyl proteom, methyl proteome, whatever other “—proteome”), glyco conjugation

    Transcriptome (micro-RNA regualtion, RNA regulation mechnanism)
    Genome (binding site locations)

    Well, now what we have to deal with those pesky protein domains now being appended and inserted into genes. How does that follow Theobald’s model of random mutation? Common design rather than a random walk seems a better explanation for the exquisite engineering of species-specific differences in the same gene.

  11. stcordova: It explains it by saying, “God did it because common descent can’t.”You’re inability to rebut the conflicting data points is evidence common descent can’t do the job.

    Whoosh, as usual. That isn’t an explanation, and you are still confusing the explanation for nested hierarchy with the explanation(s) for the origin of various features of organisms. Common descent doesn’t explain the origin of features. It explains the distribution of features in a nested hierarchy. Everything you say after this is relevant to the origin of features, and is therefore irrelevant to my question.

  12. ADDENDUM:

    I don’t mean to say Alternative Splicing only reconfigures the protein domains in a splice variant of a protein. It can re-configure the un-ordered regions as well, perhaps moreso:

    https://www.ncbi.nlm.nih.gov/pubmed/23562110

    any proteins are composed of protein domains, functional units of common descent. Multidomain forms are common in all eukaryotes making up more than half of the proteome and the evolution of novel domain architecture has been accelerated in metazoans. It is also becoming increasingly clear that alternative splicing is prevalent among vertebrates. Given that protein domains are defined as structurally, functionally and evolutionarily distinct units, one may speculate that some alternative splicing events may lead to clean excisions of protein domains, thus generating a number of different domain architectures from one gene template. However, recent findings indicate that smaller alternative splicing events, in particular in disordered regions, might be more prominent than domain architectural changes. The problem of identifying protein isoforms is, however, still not resolved. Clearly, many splice forms identified through detection of mRNA sequences appear to produce ‘nonfunctional’ proteins, such as proteins with missing internal secondary structure elements. Here, we review the state of the art methods for identification of functional isoforms and present a summary of what is known, thus far, about alternative splicing with regard to protein domain architectures.

    So what would the effect be if a spice variant doesn’t change the protein domains in a particular protein? Recall the other “-omes” I mentioned? The Phospho proteome, the Acetyl Proteome, the Methyl Proteome, the Glyco conjugations? These splice variants likely will have different behaviors for these “-omes”.

    Perhaps a tangible example. The c-terminus of Topoisomerase in humans have very much more Serine amino acids than the same Topoisomerase in other creatures. A mass-spec study I saw privately showed these were phosphorylation sites (as in the phospho proteome). That means bits of information are being flipped on and off in a way not available in other creatures with the homologs of that particular Topoisomerase.

    Now, it takes a little bit of thought, but one can see the an alternative splice with a different pattern of Serine residues results in a different pattern of phospho proteome and other regulation. Voila!

    No papers on it yet, but I can sense they should be forthcoming. That’s a testable prediction.

  13. stcordova: Depends on who you ask.I would say structurally and morphologically that there is a nested hierarchy.Creationists were among the first to note this.

    Why does this nested hierarchy exist?

    Evolutionists have claimed the nested hierarchy of structures is proof of Universal Common Ancestry, but one could also argue the nested hierarchy of structures is evidence against macro evolution.How? Mammals come from other mammals, birds from birds. Birds don’t come from fish. We should expect fish give rise only to other fish. That is a structural argument against Macro Evolution.I also showed that Cox1 agrees with that claim.

    No, you showed that you are not competent to assemble a data set or analyze molecular data. I have explained the problems with your analysis, but you ignored all that. In fact, the nested hierarchy shows that birds and mammals are amniotes, that amniotes are tetrapods, that tetrapods are sarcopterygians, that sarcopterygians are osteichthyans, and so on. How do you account for that?

    Nested hierarchies of structure leads to the chicken and egg paradox, if mammals give rise only to mammals, what gave rise to the first mammal?

    You clearly don’t understand nested hierarchy, which has groups within groups. Any descendant of a mammal is, by definition, a mammal. Any descendant of a therapsid is, by definition, a therapsid. Mammals are therapsids because they’re nested within therapsids, and we have quite a nice fossil record of the transition between non-mammalian therapsids and mammals. Have you ever looked at that?

    Of course you don’t even think mammals are related by common descent to other mammals. At times you have seemed to claim that every species is a separate kind. It’s hard to argue with you when you won’t take a position on what are and are not separate kinds, but doesn’t seem at least odd that you can’t tell?

  14. No, you showed that you are not competent to assemble a data set or analyze molecular data.

    I showed I wasn’t willing to cherry pick genes and computational methods that give you the results that agreed with your forgone conclusions. Cox1 gives a different nested hiearchy than BMP2. I demonstrated that. The distance matrices on the genes show the conflict.

  15. colewd:
    John Harshman,

    A straight forward question that you would ask.Can you define what you are claiming validates common descent?

    Bill, you can’t write clearly and you can’t tell when you aren’t writing clearly. Just take my word for it that you aren’t communicating. For a start, when you want to say something, use the right word, not its second cousin. “Define” is at best a fourth cousin.

    Now, what I am claiming validates common descent is the presence of nested hierarchical structure in phylogenetic data. We can detect that structure in a number of ways, but most of that boils down to consistency among different bits of data, all supporting one particular arrangement above others. It’s exactly the pattern we would expect if the species being compared shared a branching history. Nobody can think of another reason for this pattern, least of all you. (We also expect a certainly amount of homoplasy, some of it due to selection and some of it purely stochastic.)

  16. stcordova: I showed I wasn’t willing to cherry pick genes and computational methods that give you the results that agreed with your forgone conclusions.Cox1 gives a different nested hiearchy than BMP2.I demonstrated that.The distance matrices on the genes show the conflict.

    You haven’t shown any distance matrices, just a couple of isolated distance comparisons, none of which show what you think they show. The particular set of poorly chosen Cox I sequences you picked, over the objections of competent phylogeneticists, show a problematic result; that’s all. (Need I remind you that you originally thought they were cytochrome c? I mention this to point out your naivety and incompetence on a matter we both now agree on.) You have chosen a grand total of two genes, with conflicting results. I invite you to try a few more nuclear genes and/or a greatly expanded and well chosen taxon sample for Cox I. (Ciona is not a good idea.)

    Anyway, you believe the Cox I result and reject the BMP2 result. Why? Any competent phylogeneticist would have told you, in advance, that a nuclear gene would be preferred to a mitochondrial gene. (And of course you originally thought you were sampling a nuclear gene; the mitochondrial gene is just an accident of your misunderstanding.)

  17. John Harshman,

    It’s exactly the pattern we would expect if the species being compared shared a branching history.

    What do you mean by “sharing a branching history”?

  18. John Harshman:

    It’s exactly the pattern we would expect if the species being compared shared a branching history.

    No it’s not because Species Trees are often in conflict with gene trees, and gene trees are often in conflict with other gene trees.

    The morphological nested Hiearchies of Linnaues are nice approximations. Hennig was at least right in as much as his school isn’t just protein-centric. There is more to biology than the building blocks of proteins, but how these building blocks are put together.

    The branching pattern of morphological forms is the sort of pattern that designer would construct to resist claims of macro evolution. The nested hierarchy doesn’t suggest birds came from fish, it suggests birds came from other birds.

  19. stcordova: No it’s not because Species Trees are often in conflict with gene trees, and gene trees are often in conflict with other gene trees.

    What is the statistical significance of that conflict, and how are species trees inferred in the first place?

    Please answer both questions substantively.

  20. John Harshman,

    It’s exactly the pattern we would expect if the species being compared shared a branching history. Nobody can think of another reason for this pattern, least of all you. (We also expect a certainly amount of homoplasy, some of it due to selection and some of it purely stochastic.)

    Why would you expect to see any homoplasy?

  21. Rumraket:

    What is the statistical significance of that conflict,

    In some cases close to 100% significant. If the gene is so mangled in one species so as to be close to non-existent, it’s hard to build a gene tree.

    Try it with BMP2 (bone morphogenetic protein 2). I could hardly find homologs of it in Lungfish, but there was one in a coelecanth and human. The Mackarel (vertebrate) was more divergent than Ciona (a non-vertebrate) from humans. It could be an issue with sequencing depth. But I should point out, there is a sampling bias issue — cherry picking. I find that deeply troubling.

    We tend to build gene trees with genes that are nicely represented across taxa. Hence we can assume a gene mutates slowly over generations. But if the gene is missing in a lineage, well then it mutated so quickly it dissappeared, so how do we put that in our data to build clades? What if you have situation the only candidate homolog is something you decide is a not homolog and then you throw that data point out and move on to another gene that gives you a nicer set of homologs that agrees with your forgone conclusions?

    I showed the pan-genome (the infamous “flower diagram”) of some vertebrates. Chickens and humans shared some genes to the exclusion of other creatures. Building a gene tree would be “chickens and humans” in one clade, and all the other creatures in another. That’s statistically significant to me — as in close to 100% disagreement with all the other trees!

    Now for issues raised with the Pan-Genome diagram (like genes shared only by chickens and humans), reductive evolution is a reasonable explanation for the flower pattern, but reductive evolution isn’t constructive evolution. It creates a problem for how the pan-genome arose and how big it was. Paul Nelson, at the ICC 2013 conference, mentioned in passing it would have to be outrageously large. I don’t recall the citation.

    It’s only fair to point out, suppose we have the same exact data set on the same gene. You’ll get a different phyogeny based on the parameters you throw in and which algorithm you use and which species should provide the outgroup. I don’t find this very reassuring.

    Of course then you can appeal to consensus and ignore the outliers (like genes shared only by chickens and humans), and get as perfect looking a phylogeny as you please. But that is a cherry picking problem. That’s not evidence against some global consensus nested hierarchy like Linnaeus suggested, it’s evidence against gene trees always being consistent with species trees.

    But let’s back up from the question of common design, and just ask what is the effect of HGT. It makes a mockery of using gene trees to make species trees. And if protein domains in Eukaryotes are shuffled between proteins, you start to have comparable issues. It’s unavoidable.

    https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2873008/

    What seems indisputable is that we can identify organisms that have synapomorphies, both genetic and phenotypic. However, even though we recognize groupings, we do not have a bacterial species concept and we do not understand how these groupings (species, subspecies, even genera) form. Multi-locus sequence analysis (Gevers et al. 2005) has shown that there is some structure among currently defined species (Falush et al. 2001; Kidgell et al. 2002; Achtman & Wagner 2008, Buckee et al. 2008). However, this kind of analysis, which has been carried out extensively in thousands of isolates, has the limitation that it only examines the evolutionary history of a set of core genes. Not only does this limit the amount of information used in the analysis, core genes are not representative of the rest of the genes in a genome in terms of factors such as functional category and rate mutation. For a modern system of classification to work, it must use complete genomes and be able to accommodate HGT.

    The concept of prokaryotic species is difficult to address, and there is considerable diversity of opinion on what constitutes a species among the prokaryotes.

    You asked:

    and how are species trees inferred in the first place?

    Well if professionals can’t answer the question, I sure can’t, but if gene trees conflict, if there are species trees built from them, there will be at least on gene tree that conflicts with the species tree. Hence we can prove species trees will disagree with gene trees even if we don’t have a stable definition of a what constitutes a species tree!

  22. colewd: John Harshman,

    It’s exactly the pattern we would expect if the species being compared shared a branching history. Nobody can think of another reason for this pattern, least of all you. (We also expect a certainly amount of homoplasy, some of it due to selection and some of it purely stochastic.)

    Why would you expect to see any homoplasy?

    He just told you.

    But why should that prevent you from asking what you ought to know already when you’re acting as if you understand these matters?

    Glen Davidson

  23. colewd:
    John Harshman,

    What do you mean by “sharing a branching history”?

    How is it possible that at this late date you don’t know this? I mean what we’ve been talking about all this time, that species are related by common descent, generally with species splitting into two or more species and this repeating until there is a branching tree of life.

  24. stcordova: The branching pattern of morphological forms is the sort of pattern that designer would construct to resist claims of macro evolution. The nested hierarchy doesn’t suggest birds came from fish, it suggests birds came from other birds.

    Are you suggesting that there’s a nested hierarchy precisely because the designer didn’t want to provide evidence for macroevolution? That makes no sense, as the nested hierarchy is among the chief pieces of evidence for macroevolution. While you have finally presented a hypothesis to explain nested hierarchy, it’s the lamest possible hypothesis. And of course the nested hierarchy suggests that birds came from theropods, which came from dinosaurs, which came from archosaurs, which came from diapsids, which came from amniotes, etc. Even turtles aren’t turtles all the way down.

  25. colewd:
    John Harshman,

    Why would you expect to see any homoplasy?

    Again, I’m surprised that you don’t know this at this late date. If mutations are random and there are a limited number of possible states, we expect the same state to re-occur every so often. For example, the only possible states at a given position in a DNA sequence are A, C, G, or T. If a mutation happens to a site currently holding A, it must be C, G, or T. (It happens that the probability of G is generally higher than C or T, but that’s a complication we don’t have to think about.) Let’s just suppose a mutation has a 1/3 chance of being each of the 3 bases. The probability that mutations in separate lineages will both be C is thus 1/9.

  26. stcordova: No it’s not because Species Trees are often in conflict with gene trees, and gene trees are often in conflict with other genetrees.

    Gene trees conflict with species trees less often than you may suppose, and of course the most common reason, lineage sorting, is well known and introduces only small differences. Now what you are probably talking about would be estimated gene trees, which conflict more often, but that shouldn’t be taken as evidence that the species trees conflict unless the estimated trees have been tested for alternative explanations, as your Cox I tree has not.

    The morphological nested Hiearchies of Linnaues are nice approximations.

    Approximations of what, exactly? Why are there nested hierarchies?

  27. Perhaps it might be helpful to give an example of a Promiscuous Domain that is found in many proteins.

    https://en.wikipedia.org/wiki/Pleckstrin_homology_domain

    Pleckstrin homology domain (PH domain) is a protein domain of approximately 120 amino acids that occurs in a wide range of proteins involved in intracellular signaling or as constituents of the cytoskeleton.[1][2][3][4][5][6][7]

    This domain can bind phosphatidylinositol lipids within biological membranes (such as phosphatidylinositol (3,4,5)-trisphosphate and phosphatidylinositol (4,5)-bisphosphate),[8] and proteins such as the βγ-subunits of heterotrimeric G proteins,[9] and protein kinase C.[10] Through these interactions, PH domains play a role in recruiting proteins to different membranes, thus targeting them to appropriate cellular compartments or enabling them to interact with other components of the signal transduction pathways.

    Ok so this is a 120 amino acid motif. What proteins “inherited” this motif?

    https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4704004/

    Membrane and Protein Interactions of the Pleckstrin Homology Domain Superfamily

    The human genome encodes about 285 proteins that contain at least one annotated pleckstrin homology (PH) domain. As the first phosphoinositide binding module domain to be discovered, the PH domain recruits diverse protein architectures to cellular membranes.

    Unfortunately it is not clear in that paper how widespread the insertion of PH domain was into supposedly pre-existing genes. But the PH domain is one of the most promiscuous domains inserting itself in genes.

    Such promiscuous domains evidence of common design vs. common descent because one must appeal to rather exotic mechanisms to make the insertion (in the right place no less!) and then make the insertions functional. Beyond that, the plextrin domain when it lands promiscuously on a pre-existing protein is often coupled with another promiscuous domain. If this happened by common descent, it certainly doesn’t look random to me, otherwise this is like expecting tornados to pass through a junk yard and expect something functional to emerge.

  28. GlenDavidson: I showed the pan-genome (the infamous “flower diagram”) of some vertebrates. Chickens and humans shared some genes to the exclusion of other creatures. Building a gene tree would be “chickens and humans” in one clade, and all the other creatures in another. That’s statistically significant to me — as in close to 100% disagreement with all the other trees!

    Let’s be clear: you showed a diagram with exactly four vertebrates on it. If that diagram had sampled many more species, we would be able to get a clearer idea of exactly what had happened, and when. I would suggest that the greatest likelihood would be that most genes found in chicken and human but not mouse and zebrafish arose in sarcopterygians after the spit between sarcopterygians and actinopterygians and were lost in the rodent lineage after the split between the primate and rodent portions of Euarchontoglires. There are probably also a few that were present in the common osteichthyan ancestor but were independently lost somewhere in the rodent and actinopterygian lineages. And that losses of different genes happened in different places on the tree, from root to tip.

    More importantly, you can’t build a gene tree based on presence or absence of one gene. Aside from the fact that that isn’t what a gene tree is, you shouldn’t construct phylogeny from a single binary character. I also don’t think you know what “statistically significant” means.

    Now for issues raised with the Pan-Genome diagram (like genes shared only by chickens and humans), reductive evolution is a reasonable explanation for the flower pattern, but reductive evolution isn’t constructive evolution. It creates a problem for how the pan-genome arose and how big it was. Paul Nelson, at the ICC 2013 conference, mentioned in passing it would have to be outrageously large. I don’t recall the citation.

    I don’t think you can say that at all. There is no reason to suppose that a gene present in one, two, or even three species in your diagram, depending on which species, was present in the common ancestor of all four. As I have pointed out, the simplest scenario for the genes shared among chicken and human is not two losses but a gain and a loss. So we have as much evidence from that diagram for constructive evolution as for reductive evolution. And again, a bigger taxon sample would clarify.

    It’s only fair to point out, suppose we have the same exact data set on the same gene. You’ll get a different phyogeny based on the parameters you throw in and which algorithm you use and which species should provide the outgroup. I don’t find this very reassuring.

    You say that as if the choice of parameters, algorithm, character sample, and taxon sample were arbitrary. They are not, and a great deal of literature concerns when various methods and samples are appropriate and when they are not. Scientists are not as ignorant as you suppose based on your own ignorance.

    Of course then you can appeal to consensus and ignore the outliers (like genes shared only by chickens and humans), and get as perfect looking a phylogeny as you please. But that is a cherry picking problem. That’s not evidence against some global consensus nested hierarchy like Linnaeus suggested, it’s evidence against gene trees always being consistent with species trees.

    Once more, you seem not to know what “gene tree” means. What you discuss is not a gene tree; more importantly, it’s evidence only that genes appear and disappear. One can construct a phylogeny based on presence/absence of genes, and I would expect it to match the standard one using any reasonable method. I bet simple parsimony would work. The data in your diagram certainly support the usual tree. That’s why the patterns that require only one change on the standard tree have many more genes than the patterns that require two changes. What’s your explanation for that?

    But let’s back up from the question of common design, and just ask what is the effect of HGT. It makes a mockery of using gene trees to make species trees.

    Still not sure you know what a gene tree is, but this time you’re right, if there’s enough HGT compared to vertical transmission. But that doesn’t seem generally true even for prokaryotes, which seem to practice HGT mostly among close relatives. And eukaryotes are considerably less prone to such things. Anyway, I thought you accepted a nested hierarchy of life. Now you’re arguing that it doesn’t exist?

    And if protein domains in Eukaryotes are shuffled between proteins, you start to have comparable issues. It’s unavoidable.

    Not if it’s rare enough.

    Let’s forget bacteria for a while and concentrate on vertebrates, where the greatest interest lies. There seems to be little horizontal transfer or domain shuffling within vertebrates. Would you agree? If there were too much it would eliminate the nested hierarchy, but you have previously agreed that this hierarchy exists. Given that it exists, it needs an explanation. What is your explanation?

  29. stcordova: In some cases close to 100% significant.

    “100% significant”. What the fuck are you talking about?

    If the gene is so mangled in one species so as to be close to non-existent

    What does that even mean?

    …it’s hard to build a gene tree. Try it with BMP2 (bone morphogenetic protein 2). I could hardly find homologs of it in Lungfish

    What does “hardly” mean, that you found some, or none at all? And what gene-tree do you get from this gene, and how does it compare to other gene trees made from other nuclear genes?

    What is the level of congruence?

    but there was one in a coelecanth and human. The Mackarel (vertebrate) was more divergent than Ciona (a non-vertebrate) from humans. It could be an issue with sequencing depth.But I should point out, there is a sampling bias issue — cherry picking.

    Sampling bias isn’t cherry picking. Those are two very different things.

    I find that deeply troubling.

    I find your flailing and incompetent use of terms you don’t understand troubling.

    We tend to build gene trees with genes that are nicely represented across taxa.

    No shit? What the hell else would you do it with?

    Hence we can assume a gene mutates slowly over generations. But if the gene is missing in a lineage, well then it mutated so quickly it dissappeared

    Is that the only possible option? And even if that is what happened, why is that a problem for the inference of common descent? Must all genes be present and mutate at a similar rate in order to infer a common genealogical relationship? If you’re going to say yes, explain why they must.

    so how do we put that in our data to build clades?

    This doesn’t even make sense. Why must we put that “in our data to build clades”?

    What?

    Seriously. Despite calls for you to try to make more sense and speak more clearly, you’re suddenly even less coherent than usual.

    What if you have situation the only candidate homolog is something you decide is a not homolog and then you throw that data point out and move on to another gene that gives you a nicer set of homologs that agrees with your forgone conclusions?

    But Has Anyone Really Been Far Even as Decided to Use Even Go Want to do Look More Like?.

    You’re saying that there’s a hypothetical situation where we only have one candidate for a homologous gene. Okay. But then we decide we don’t believe the genes really are homologous. Okay. But then… we suddenly move on to another gene, which we do believe is a nice set of homologous genes. But… wait, didn’t you start this hypothetical situation by saying that we don’t have any other putative homologus genes?

    Seriously, did you leave some sort of chat-bot with a database of random phylogenetics terms, to continue with keeping up appearances while you went out here?

    I showed the pan-genome (the infamous “flower diagram”) of some vertebrates. Chickens and humans shared some genes to the exclusion of other creatures. Building a gene tree would be “chickens and humans” in one clade, and all the other creatures in another. That’s statistically significant to me — as in close to 100% disagreement with all the other trees!

    I can only facepalm. You can’t compare the branching order of trees if the genes aren’t present in all the species used in all trees. Particularly not with only four taxons. LOL. ffs.

  30. Ah from the Wiki link! The list of proteins which Promiscuous Plextrin is present in. It’s shocking what sort of diverse places where Promisuous Plextrin shows herself.

    Proteins reported to contain PH domains belong to the following families:

    Pleckstrin, the protein where this domain was first detected, is the major substrate of protein kinase C in platelets. Pleckstrin is one of the rare proteins to contain two PH domains.

    Ser/Thr protein kinases such as the Akt/Rac family, the beta-adrenergic receptor kinases, the mu isoform of PKC and the trypanosomal NrkA family.

    Tyrosine protein kinases belonging to the Btk/Itk/Tec subfamily.

    Insulin receptor substrate 1 (IRS-1).

    Regulators of small G-proteins like guanine nucleotide releasing factor GNRP
    (Ras-GRF) (which contains 2 PH domains), guanine nucleotide exchange proteins like vav, dbl, SoS and S. cerevisiae CDC24, GTPase activating proteins like rasGAP and BEM2/IPL2, and the human break point cluster protein bcr.

    Cytoskeletal proteins such as dynamin (see InterPro: IPR001401),

    Caenorhabditis elegans kinesin-like protein unc-104 (see InterPro: IPR001752), spectrin beta-chain, syntrophin (2 PH domains), and S. cerevisiae nuclear migration protein NUM1.

    Mammalian phosphatidylinositol-specific phospholipase C (PI-PLC) (see InterPro: IPR000909) isoforms gamma and delta. Isoform gamma contains two PH domains, the second one is split into two parts separated by about 400 residues.

    Oxysterol-binding proteins OSBP, S. cerevisiae OSH1 and YHR073w.
    Mouse protein citron, a putative rho/rac effector that binds to the GTP-bound forms of rho and rac.

    Several S. cerevisiae proteins involved in cell cycle regulation and bud formation like BEM2, BEM3, BUD4 and the BEM1-binding proteins BOI2 (BEB1) and BOI1 (BOB1).

    C. elegans protein MIG-10.

    Ceramide kinase, a lipid kinase that phosphorylates ceramides to ceramide-1-phosphate.[13]

    ….
    Human genes encoding proteins containing this domain include:
    ABR, ADRBK1, ADRBK2, AFAP, AFAP1, AFAP1L1, AFAP1L2, AKAP13, AKT1, AKT2, AKT3, ANLN, APBB1IP, APPL1, APPL2, ARHGAP10, ARHGAP12, ARHGAP15, ARHGAP21, ARHGAP22, ARHGAP23, ARHGAP24, ARHGAP25, ARHGAP26, ARHGAP27, ARHGAP9, ARHGEF16, ARHGEF18, ARHGEF19, ARHGEF2, ARHGEF3, ARHGEF4, ARHGEF5, ARHGEF6, ARHGEF7, ARHGEF9, ASEF2,
    BMX, BTK,
    C20orf42, C9orf100, CADPS, CADPS2, CDC42BPA, CDC42BPB, CDC42BPG, CENTA1, CENTA2, CENTB1, CENTB2, CENTB5, CENTD1, CENTD2, CENTD3, CENTG1, CENTG2, CENTG3, CIT, CNKSR1, CNKSR2, COL4A3BP, CTGLF1, CTGLF2, CTGLF3, * CTGLF4, CTGLF5, CTGLF6,
    DAB2IP, DAPP1, DDEF1, DDEF2, DDEFL1, DEF6, DEPDC2, DGKD, DGKH, DGKK, DNM1, DNM2, DNM3, DOCK10, DOCK11, DOCK9, DOK1, DOK2, DOK3, DOK4, DOK5, DOK6, DTGCU2,
    EXOC8,
    FAM109A, FAM109B, FARP1, FARP2, FGD1, FGD2, FGD3, FGD4, FGD5, FGD6,
    GAB1, GAB2, GAB3, GAB4, GRB10, GRB14, GRB7,
    IRS1, IRS2, IRS4, ITK, ITSN1, ITSN2,
    KALRN, KIF1A, KIF1B, KIF1Bbeta,
    MCF2, MCF2L, MCF2L2, MRIP, MYO10,
    NET1, NGEF,
    OBPH1, OBSCN, OPHN1, OSBP, OSBP2, OSBPL10, OSBPL11, OSBPL3, OSBPL5, OSBPL6, OSBPL7, OSBPL8, OSBPL9,
    PHLDA2, PHLDA3, PHLDB1, PHLDB2, PHLPP, PIP3-E, PLCD1, PLCD4, PLCG1, PLCG2, PLCH1, PLCH2, PLCL1, PLCL2, PLD1, PLD2, PLEK, PLEK2, PLEKHA1, PLEKHA2, PLEKHA3, PLEKHA4, PLEKHA5, PLEKHA6, PLEKHA7, PLEKHA8, PLEKHB1, PLEKHB2, PLEKHC1, PLEKHF1, PLEKHF2, PLEKHG1, PLEKHG2, PLEKHG3, PLEKHG4, PLEKHG5, PLEKHG6, PLEKHH1, PLEKHH2, PLEKHH3, PLEKHJ1, PLEKHK1, PLEKHM1, PLEKHM2, PLEKHO1, PLEKHQ1, PREX1, PRKCN, PRKD1, PRKD2, PRKD3, PSCD1, PSCD2, PSCD3, PSCD4, PSD, PSD2, PSD3, PSD4, RALGPS1, RALGPS2, RAPH1,
    RASA1, RASA2, RASA3, RASA4, RASAL1, RASGRF1, RGNEF, ROCK1, ROCK2, RTKN,
    SBF1, SBF2, SCAP2, SGEF, SH2B, SH2B1, SH2B2, SH2B3, SH3BP2, SKAP1, SKAP2, SNTA1, SNTB1, SNTB2, SOS1, SOS2, SPATA13, SPNB4, SPTBN1, SPTBN2, SPTBN4, SPTBN5, STAP1, SWAP70, SYNGAP1,
    TBC1D2, TEC, TIAM1, TRIO, TRIOBP, TYL,
    URP1, URP2,
    VAV1, VAV2, VAV3, VEPH1

  31. stcordova: God did it because common descent can’t.

    LOL. Common descent literally predicts congruent independent trees.

    I guess when all else fails you can just brainlessly assert a conclusion diametrically opposite to demonstrable fact.

    I guess in a way you are just affirming the typical creationist doctrine that when reality and scripture conflicts, reality is wrong and scripture is right. You’ve truly taken that to heart and are following in the footsteps of Henry Morris, who wrote:
    “…the main reason for insisting on the universal Flood as a fact of history and as the primary vehicle for geological interpretation is that God’s Word plainly teaches it! No geologic difficulties, real or imagined, can be allowed to take precedence over the clear statements and necessary inferences of Scripture.
    – Henry Morris, Biblical Cosmology & Modern Science, pp 32-33 (1970)

  32. Rumraket:

    LOL. Common descent literally predicts congruent independent trees.

    But Common Descent doesn’t predict evolution of novel features that will define groups. As John Harshman himself said:

    Common descent doesn’t explain the origin of features.

    John Harshman
    TSZ, Common Descent Common Design, 11/30/17

    Origin of novel features, as in features without ancestors, is evidence against Universal Common Ancestry.

    I guess if you want to use the word Synapomorphy, you can, but a rose is a rose by any other name.

  33. Rumraket:

    You’ve truly taken that to heart and are following in the footsteps of Henry Morris

    Actually I don’t agree with Henry Morris’s approach. No disrespect to Henry Morris, God rest his soul.

    I’ve provided data points suggestive of features with no ancestor.

    Features with no ancestors is evidence against Universal Common Descent. Therefore the patterns of similarity are better explained by common design since the patterns of unique complex features are not explained by common descent.

    As our resident phylogenetict John Harshman said:

    Common descent doesn’t explain the origin of features.

    John Harshman, expert phylogeneticst and evolutionary biologist from the prestigous University of Chicago

  34. John Harshman,

    Again, I’m surprised that you don’t know this at this late date. If mutations are random and there are a limited number of possible states, we expect the same state to re-occur every so often. For example, the only possible states at a given position in a DNA sequence are A, C, G, or T. If a mutation happens to a site currently holding A, it must be C, G, or T. (It happens that the probability of G is generally higher than C or T, but that’s a complication we don’t have to think about.) Let’s just suppose a mutation has a 1/3 chance of being each of the 3 bases. The probability that mutations in separate lineages will both be C is thus 1/9.

    Are you claiming that homoplasy we observe is limited to single mutations? What is the chance that the site will mutate in both lineages? In humans there are 3.2 billion nucleotides. I think the chance of a mutation occurring simultaneously in two separate species that does not occur in a common ancestor is vanishingly small. If an adaption which occurs in two species and not in their common ancestor takes over 3 mutations then I think it falsifies the sharing of a common ancestor between the two selected species.

  35. stcordova: Origin of novel features, as in features without ancestors, is evidence against Universal Common Ancestry.

    lol, no it isn’t. Evolution explains the origin of novel features. Common descent explains the branching pattern you get from their distribution in the diversity of life.

  36. colewd: If an adaption which occurs in two species and not in their common ancestor takes over 3 mutations then I think it falsifies the sharing of a common ancestor between the two selected species.

    Why?

Leave a Reply