Evolution vs Baraminology: the empirical showdown.

Post Reply New Topic

Indeed, I don't belong to that group.

I'll let you decide what the cuttoff is for genetic discontinuity, but I was wondering just today this:

Supposing that extraterrestrial aliens had DNA and that they had no relation to us, then would the software currently available notice/detect such unrelatedness or would it generate a cladogram regardless?

Well, if it looked similar to ours then the software would still produce a cladogram. However, here is where it will fail: I can produce separate cladograms from different genes. If two species are truly unrelated, then there should be no genome-wide correspondence between them, meaning that the cladograms produced will be inconsistent from gene to gene, in a manner not explainable by things like statistical artifacts, base-compositional biases, etc.

So basically, what I am looking for as a demonstration of genetic discontinuity is for phylogenetics to become incoherent when jumping from one "kind" to another. I think that's a fair prediction to make.

Woot! My university just opened access to our new supercomputer. That should work nicely to complement the physics cluster for the calculations we'll need to do.

Ok. So what you are saying is that if you test two sight genes from parrots against two matching genes from parakeets, both genes should show a similar amount of change as both genes evolved over a similar time period. If you were to represent the number of differences as pieces of string, genes A and B should be of similar length. (but its represented as a tree in this case)

If baraminology holds however, then the strings might well be random lengths. That is, the trees will likely differ.

Ambiguity can be eliminated by testing a third gene.

If that is what you mean, it is similar to how my family is trying to connect with other families of the same surname.

In regard to what I think regarding the delineation of kinds, at least based on morphology, parrots would all be of the same kind, including sub-members of parakeets, cockatiels, conures, African Grays, Macaws, Cockatoos, Galahs, Keas, et al. However, ducks would not be of the same kind; they are a different kind. As would be sparrows, finches, canaries; various varieties of penguins; turkeys; chickens; pigeons & doves; etcera. If you were to replace the word "parakeet" with a bird which does not belong to the parrot kind, then what you said would be more corresponding with where I view the distinction of created kinds. If I misinterpreted what you said though, I'm sorry. Rather tired here, since I've been staying up all night in order to help my wife wake up for work at 4:30 in the morning (shift starts at 6am, but she prefers to be woken up more gradually) and also help her get ready in the morning, so I am rather tired. But parakeets are a member of the parrot kind, and provided that information would be lost rather than gained over time, the archaebaramin would have had motile crests as cockatiels do, but also they would have had more coloration (or at least the information in the genome for the coloration to be distributed among the various sub-members of the parrot kind in the patterns they have today).

Sort of. We are not going to be looking directly at the number of changes though, but rather letting computer algorithms infer relationships from those differences. Some of the algorithms will output the length of branches as corresponding to the number of changes since a proposed common ancestor, but the topology of the tree is the only really interesting thing here. If baraminology is correct, the trees produced by genes A and B should not generally be topologically equivalent.

And we'll be testing a lot more than three genes. I did an example problem last week with 20 genes, and that seems like a decent number to look at for this problem. We have to make sure that the genes are unlinked and preferably balanced between coding and non-coding.

Any objections to my sequence alignment procedure? I have found some other programs that might do a better job than Clustal (I have been corresponding with a prof who tends to regard Clustal as somewhat unreliable) and I have some documentation to read before I do any more.

I don't know that I should have any objections to your procedure. However, in regard to the balance between coding and non-coding genes, I had read that the non-coding genes generally have a high degree of similarity and, from a design perspective, it may be like code to provide syntax and context for how the coding genes are to be employed. If you want to go with some homeobox genes along with protein coding genes, that would probably be okay. IDK though, I'll defer to you.

Good point. I should have used broader examples.

So keets and parrots can be crossed? Interesting.

Well, as far as hybridization goes it is like this:

If two species can hybridize and produce viable offspring (whether fertile or not), then they are of the same "kind"/baramin.

I phrase it in that manner because hybridization does entail that form of relation, but lack of hybridization is not necessarily an indication of non-relatedness.

Well, one thing is that protein-coding genes are more likely to be constrained by convergent evolution. Two species in the same environment will develop similar structures. Non-coding strands of DNA are not as likely to be directly affected by natural selection, meaning that we just have plain genetic drift there in many situations. That gives us a more neutral marker of what changes we have seen since two species split from a common ancestor.

From the design standpoint, if the non-coding genes are just syntax and/or context, then they may as well just be roughly the same across all species, regardless of kind. That works perfectly for this experiment, because if there is no connection between relatedness of two species and the similarity of their non-coding DNA, then analysis of non-coding DNA should give me ridiculous phylogenies as often as it gives plausible-looking ones. For instance, I might get parakeets placed within the Passer clade from non-coding genes.

As for "Two species in the same environment" members of the parrot kind are literally all over the world. The Australian outback, Australian jungles, Central American jungles, South American jungles, New Zealand plains and forests, African jungles, and there was even an indigenous population on the eastern seaboard of North America before settlers from Britain arrived. They tend to prefer warmer weather, so in that sense they are in the "same" basic environment, however their distribution practically forbids the environments from being exactly equivalent. If you wish to use members of the parrot family from various continents, that may help somewhat. Your option though, based on what is available and what you think best.

Not all non protein coding ones would just be for syntax or context though. Proteins provide the shape of components to be utilized, but there still needs to be a blueprint as to where such materials are used and in which order they are assembled. The homeobox genes are some of such genes that I know about which provide a basic large-scale anatomical ordering of "This goes here with respect to this", but there should also be genes which provide such data with regard to the structure and order of assembly of organelles and placement of various types of tissues within multicellular organisms.

I'm not sure what you are meaning by non protein coding genes. There are genes that code for regulatory RNA molecules, but these will likely be just as conserved as protein coding genes. Are you meaning the regulatory sequences (or untranslated regions) of protein coding genes, or the non-coding DNA between genes? If the former, these are often not well defined (apart from UTRs), and if they are well defined it is because of conserved sequences, so you will not have the independence that you wish. If the latter, then how can you be sure to pick equivalent sequences in the species studied? You can look for regions between two conserved genes (where gene order is also conserved in all the species studied), but there is always the possibility of insertion of transposable elements or other pieces of DNA, which will skew results (although should be recognisable with alignment programmes). Homeobox genes are protein coding - they are transcription factors that bind DNA to regulate the transcription of other genes and are extremely highly conserved. Protein coding genes also provide the information as to the positioning of components within the cell (this is partly what I am working on), and often these are also highly conserved between distantly related species.

slight thread hijack:

Ok, a quick search of the KJV turns up entries for horses and asses (sometimes both in the same passage (eg, Gen. 47:17), so it's not just translating the same word different ways). Horses and Asses can interbreed and produce mules (also referenced in the KJV, eg. Gen. 36:24). Well, you might say, mules are sterile: therefore, different 'kinds.' Yes?

No.
Turns out, mules aren't always sterile. Female mules ('mollies') can and do produce offspring from either jacks or stallions.
http://en.wikipedia.org/wiki/Mule

So: Asses and horses, listed as separate species in the bible, but somewhat cross-fertile. Same type, or not?

You're going by the criteria for species firstly, which is that the offspring need to be fertile. The criteria for being in the same kind, at least from within a decade ago, is that the offspring produce viable offspring, meaning alive, whether they are fertile or not. Horses and donkeys would be of the same kind just due to producing viable offspring, the mules, whether or not the mules could reproduce. As for your selection of Bible passage, the time of the events of Genesis 13 to 50 would be not be at the time of Creation or the Flood, but rather centuries, and at least hundreds of generations, after. It may have been that horses and donkeys diverged prior to the Flood, but I don't know.

Ok, so you're saying that horses and asses are the same 'kind' but managed to evolve enough differences in 6000 years (or probably significantly less time than even that) to be counted in modern biology as different species? That's pretty darn fast.

And also that separate descriptions in the bible does not necessarily mean that the two organisms are different 'kinds.'