Uncommon Descent Serving The Intelligent Design Community

“Punctuated Evolution”


In this week’s Nature, an analysis of the human genome has shown tracing 4,692 homologous recombinations backwards in time leads to “24 distinct groups”. They say their work supports a “‘punctuated’ model of evolution.” I don’t have access to the entire article (maybe somebody does), but the language contained in the abstract is the kind that we might associate with “front-loading”; viz., “Our analysis reveals that human segmental duplications are frequently organized around ‘core’ duplicons, which are enriched for transcripts and, in some cases, encode primate-specific genes undergoing positive selection.”

I did a Wikipedia search for “duplicon” and found nothing; but a Google search gives the following article: Abstract Only. Duplicons appear to be something that is seen at the chromosomal level, and involves low-level repeats of the chromosome. The description this week’s Nature authors give of a duplicon “enriched for transcripts” seems to strongly suggest that the rearrangements that have taken place in the “duplicon” have resulted in more of the genome being expressed. The image I have of what might be taking place comes from my very ancient familiarity with computer programming, and, assuming programming essentials haven’t changed much, it is this: in a computer program there are decision nodes and “go to” nodes that redirect the program to various subroutines, these subroutines being present at some numbered location along the length of the program. If for some reason one subroutine in the program were substituted for another, the program would probably still run, but the output would certainly be different. And, if additional “go to” nodes were “copied”, more subroutines would be expressed. Likewise, if you have genetic instructions along the string of nucleotides that redirects the genetic program to some other downstream part of the genome that allows some particular protein/regulatory function to take place, then, through recombination, different “subroutines” might be inserted, or more signals for transcription might be included.

Maybe this is straining the programming analogy, I don’t know. But in any event, what the authors are reporting doesn’t sound to me like information is being generated (gradualism), but that already present information is being more robustly used (punctuated model of evolution).

Any thoughts?

Human segmental duplications are hotspots for nonallelic homologous recombination leading to genomic disorders, copy-number polymorphisms and gene and transcript innovations. The complex structure and history of these regions have precluded a global evolutionary analysis. Combining a modified A-Bruijn graph algorithm with comparative genome sequence data, we identify the origin of 4,692 ancestral duplication loci and use these to cluster 437 complex duplication blocks into 24 distinct groups. The sequence-divergence data between ancestral-derivative pairs and a comparison with the chimpanzee and macaque genome support a ‘punctuated’ model of evolution. Our analysis reveals that human segmental duplications are frequently organized around ‘core’ duplicons, which are enriched for transcripts and, in some cases, encode primate-specific genes undergoing positive selection. We hypothesize that the rapid expansion and fixation of some intrachromosomal segmental duplications during great-ape evolution has been due to the selective advantage conferred by these genes and transcripts embedded within these core duplications.

I'm adding this link to an article that appeared in Nov 12's PhysOrg.com. Here it is:http://www.physorg.com/news114266805.html The topic is ERV's. Since I first ran into ERV's, my sense was that we are dealing with identification points along the DNA strand---addresses, if you will. I think this ties in quite nicely with my analogy to a computer program. Though someone has written that the program, with shifted subroutines, would fall apart, then why not ask the question, but what if the subroutine has its own address? And what if the "program" has the ability to send the "address/subroutineunits" to various parts of the DNA that are, themselves, "pre-programmed" to receive such "address/subroutine units"? My prediction, based on the ID-based assumption that we're dealing with an integrated information system that has been designed is this: ultimately, ERV's will be found to be a kind of USB of the DNA, where all sorts of different "A/S units" can be shunted in and out, with the concomitant result that different genomic "configurations" will be produced by this process. My prediction. Let's see what happens. And what would neo-Darwinism's prediction be? There wouldn't be one. Since neo-Darwinism cannot, by definition, deal with purpose/intent/design, it similarly cannot predict. That's right-----it's unfalsifiable. PaV
Co-evolution -- the mutual evolution of co-dependent organisms, e.g., bees and flowering plants -- is a dilemma for both gradual and punctuated evolution. In co-evolution, unlike evolutionary adaptation to fixed widespread physical features of the environment, e.g., water, land, air, and climate, there may be nothing to adapt to because the corresponding co-dependent trait in the other organism may be initially absent. Larry Fafarman
Here I've found some information about duplicons http://alpha01.dm.unito.it/personalpages/cerruti/mathnews.html kairos
If you did those sort of changes to a program, the program would probably crash, not produce different output. WinstonEwert

Leave a Reply