JAM
Posts: 517
Joined: July 2007
|
[quote=Daniel Smith,Oct. 08 2007,02:18] [quote=JAM,Oct. 07 2007,13:27] [quote=Daniel Smith,Oct. 07 2007,05:46]By this definition, I'd say that anything that is transcribed would qualify as functional - since the cellular machinery is going through the trouble of transcribing it.[/quote]
That is a prediction of an intelligent design hypothesis, but MET (non-Darwinian) predicts that there will be loads of RNA that has no function.[/quote]
Then this is what we need to test.[/quote]
OK, what do you propose? How about "knocking out" regions using the technology whose developers won the Nobel Prize today?
[quote] | Quote | | Quote | | So this would include protein coding sequences as well as ncRNA sequences, and anything else that's transcribed. |
But if we find anything that's transcribed but not functional, your hypothesis is dead, correct? |
As for my hypothesis being "dead" if we find anything that conflicts with what I've predicted: I don't really think that's fair since scientists are constantly finding things they don't expect and simply adjust their hypotheses to fit the evidence when they do.[/quote]
That's only true for hypotheses that have a track record. Yours doesn't. I've trashed many hypotheses that I've endorsed in previous publications, for example.
| Quote | | I will not therefore totally abandon my hypothesis if the results are different, I will simply adjust it (unless the results completely shoot it out of the water). |
It's your ethical responsibility to state the results that would completely shoot it out of the water.
| Quote | | Quote | | Quote | | I also must clarify that I do actually believe that all functional sequences (as I've defined them) are evolutionarily constrained. It's just that I don't think you can find functionality or constraint by comparing sequences to other lineages (since I posit that there are no truly neutral sites). |
What if some sites have far greater rates of change over time, Daniel? |
This is precisely the issue. How do we know the rate if it turns out that there are no neutral sites? |
Easily and empirically. We know the mutation rate in the absence of selection. If the rate of change over time is the same as the mutation rate, the only rational inference is the absence of selection. If it is less than the mutation rate, we infer selection.
[quote] | Quote | We must first determine that these sites are truly neutral and are actually accumulating mutations.
| Quote |
What sequences are used for forensic DNA analysis? |
That's a tough question, and I'm not sure I know the best answer for that. |
I'm not asking for the "best answer," I'm simply asking for an answer. Certain sequences are used for forensic DNA analysis. What are their characteristics? How polymorphic are they?
| Quote | | Quote | | Quote | | If comparing to other lineages, the function must first be known and then the entire sequence that provides that function compared. |
Not a problem.
| Quote | | However, the only true test of constraint is comparison to ancestral DNA within the same lineage. |
Oh-oh...it looks like I'm going to have to retract my retraction. Your prediction: | Quote | | there are many functional sequences that are different (even radically so) amongst related lineages - this due to their being of designed, not mutational, origin. |
makes clear predictions about the relationships between modern sequences. No ancestral sequences are required. |
I'm not backing off my original prediction, but I think certain terms mean different things to both of us, so I'm just trying to clarify. |
Good. Please define "lineage" for starters. Are mice and humans in the same or in different lineages?
| Quote | | I believe that most (if not all) sequences in a genome are functional and therefore resistive to mutation (constrained). This means there are no neutral sites that are accumulating mutations. |
And the evidence shows that you are wrong.
| Quote | | I also believe that macroevolution (when it happens) is not the result of accumulating mutations but is rather; saltational - that is - it creates new types that may have sequences that are radically different from the sequences from which they diverged (hence my earlier prediction). |
Interesting. Of the ~30,000 protein-encoding genes in the mouse and human genomes, how many do you believe/predict are absent in mouse and present in human, and vice versa?
| Quote | Therefore, this is what I expect:
1. Sequence comparisons between related lineages will result in a mixture of like and unlike functional sequences. |
I need a rigorous definition of "lineage" before pursuing this one.
| Quote | | 2. Sequence comparisons within the same lineage will show evolutionary constraint across the board - even in what are presently considered neutral sites. |
This is trivially easy to do online. Are you interested or afraid to do so?
| Quote | | 3. What are presently considered neutral sites will be found to be "instructional" - that is, they will carry the instructions that tell the various proteins, RNA and enzymes where to go, when to go and what to do when they get there. |
a) How does that relate to today's Nobel Prize?
b) How about classical genetics--do any homozygous normal inversions exist? If so, doesn't that mean that the sequences disrupted by both breakpoints have no function?
| Quote | | Now, the third prediction is more of a guess, but I think it makes sense. We know about sequences that code for proteins, and we know about sequences that regulate them, but we don't know how a certain protein "knows" where to go, what to do and when to do it. |
We know a lot about that, Daniel. In fact, it's what I've been working on for the last 16 years, since mouse genetics dumped me into the field.
I'll give you a taste--nothing about the mechanisms involved suggests intelligent design.
| Quote | My guess is that these instructions are carried in what are presently considered neutral sites and - for that reason - these sites resist mutations just like all other evolutionarily constrained sites.
I hope that's clearer. |
Except for your definition of "lineage," yes. As an introduction to how proteins "know" where to go, you might want to Google "signal sequence" and "nuclear import."
Those are the simple signals intrinsic to the protein. What I study is an order of magnitude or two more complex, fluid, and fuzzy.
|
