Already covered in this thread.
Thanks. @moderators can close this thread if they want. I had even skimmed through that thread you linked but without reading the news article. So when I came across it today I thought it was something new. As you can see by the title I gave the thread, I thought the function here seemed kind of surprising. Maybe thatâs because I donât know much, or donât quite understand it? But perhaps itâs of a kind that would maybe even surprise creationists?
But I donât knowâŚ
Merged with the other thread, and reopened for new comments.
I wonder about creationists.
On the one hand, they donât like the idea of junk DNA and would like to show that it has a function.
On the other hand, in order to see this as a function they would have to concede that speciation happens.
Personally, I donât find this particularly surprising. For several years, I have been tentatively assuming that the junk DNA might play a role in evolution. Itâs a database of genes that have been useful in the past and possibly could become useful again in the future. And it could be seen as a laboratory for experimenting with recombinant DNA while being protected from selection.
Okay, in terms of the creationists trying to use this to say âHA! See? No junk DNA!â, three points:
-
The actual part of the genome in question here is <<<10%. so good luck with the âjunk doesnât existâ argument based on this alone.
-
The question, in terms of functionality, is whether the sequences in question are conserved relative to protein coding sequences. If so, functional. If not, almost certainly not.
-
The authors specifically describe how the thing theyâre describing happens in one species of drosophila but not others. So the satellites in question in other species donât have the function described here, which indicates a secondarily acquired function for generally nonfunctional DNA. Which supports the notion that âjunkâ DNA (i.e DNA sequences without selected functions) exists.
Both of those hypotheses fail the onion test.
Yes, Iâve noted that.
They donât concede this? Or you mean just in the way the paper suggests?
I looked up satellite DNA on search creation.org just because I was curious what creationists were saying about it. The only result I remember at the moment was Tomkins saying that chromosome fusion always involved satellite DNA and there is none in the supposed human chromosomes fusion. Hopefully Iâm explaining that right. But interesting.
These points seem to be begging the question to me - why would creationists assume the evolutionary model?
Regarding your 3rd point - why couldnât it be that the function "kicks in* only when needed?
This is what I thought was interesting from a creationist perspective, if Iâm understanding it right:
When the researchers deleted a protein called Prod that binds to a specific satellite DNA sequence in the fruit fly Drosophila melanogaster, the fliesâ chromosomes scattered outside of the nucleus into tiny globs of cellular material called micronuclei, and the flies died. âBut we realized at this point that this [piece of] satellite DNA that was bound by the Prod protein was completely missing in the nearest relatives of Drosophila melanogaster,â Jagannathan said. âIt completely doesnât exist. So thatâs an interesting little problem.â
How did any flies survive if they lost that sequence and they are all part of one original baramin?
If that piece of satellite DNA was essential for survival in one species but missing from another, it could imply that the two species of flies had evolved different satellite DNA sequences for the same role over time.
This seems like the sort of evolution that creationists say doesnât exist. Thatâs what piqued my interest. How else to explain it? Perhaps thatâs what @nwrickert meant.
Welp, this has gotten over to Reddit. Some people think that the junk portion not being conserved is somehow evidence against common ancestry, even though that is exactly what we would expect if the sequences are junk.
If thatâs what creationists will argue, then weâre back to âokay, provide evidence of the creation in that stateâ, and that ends up as a âhere are the hallmarks of designâ argument, one of which isâŚlack of junk DNA! So thatâs a circular argument. Itâs a âwell god just made it that wayâ argument in which any observation is compatible with creation. Which is the case! But itâs not a good argument.
I suppose that could be so, but in order to be conserved that function must âkick inâ often enough to make a difference in reproductive fitness. Itâs a use it or lose it situation.
My educated guess is a species would need many generations of ânot usingâ that function before accumulated mutations make it non-functional.
You should only expect this if the observed/measured divergence fit into the population genetics model given an observed mutation rate and time.
Thatâs not the argument - the argument is the comparison to the evolutionary model and predictions.
I was just reading about satellite DNA and the article was saying that because it has so many repeats it accumulates a lot of mutations and therefore isnât affected by selection (because no organism has am advantage here?)
So to me, it looks like something very cool is going on here, like thereâs backups to backups to backups. If it loses a key it has something else to use? And since selection doesnât see it it means species can diversify without going instinct?
No idea if Iâm actually making a coherent argument. Lol, Iâm not getting normal sleep. Boys at grandmaâs and baby wants to sleep on me so this interesting to think about since I donât have anything better to do.
Anyway, the genome is pretty awesome. Just like cosmology, wait 20 years - the science is going to change a lot. Exciting!
What does this mean? Anyone?
I tried. Couldnât figure it out.
no positive advantage, no positive selection. a harmful mutation could be selected against.
It is WAY cool!
This topic was automatically closed 4 days after the last reply. New replies are no longer allowed.