Evolution doesn't make junk

[nancylebov]

http://www.newscientist.com/article.ns?id=mg19426086.000&feedId=online-news_rss20

A mammoth investigation of human DNA finds it generates far more RNA than thought - so what is the excess for?

Comments

[dr-zrfq.livejournal.com]

<smartass>
Where'd they get the mammoths to do this investigation?
</smartass>

[rysmiel.livejournal.com]

<deadpan>
There are estimated to be about a quarter of a million of them buried in permafrost in Siberia, or at least there were a few years ago, I've not seen estimates of how global warming may have affected this. That's enough for a lot of studying.
</deadpan>

[nancylebov.livejournal.com]

Mammoths are not junk. Mammoths are massively cool. I'm not sure I'd ignore a recreated dinosaur (a big one) to look at a recreated mammoth, but I just might.

[rysmiel.livejournal.com]

Taking this with the cathedral-sized grain of salt that New Scientist seems to require more and more these days, and not having read the Nature article and open to correction from anyone who has. one thing I would suspect about the supposed "junk" is that some of it is parasitic elements transcribed to RNA and then back to DNA somewhere else - something like a retrovirus, but without the bit where it generates a viral coat and exists outside the cell. There's a great deal we don't understand about transcription switches and control elements in general, but lots of them appear to come from fairly small elements and weak patterns, so I can see that sort of thing arising by random mutation in a "junk" sequence fairly easily.

Evolution's not just operating at the scale of the whole organism, it is also operating at the scale of replicating entities within the genome, and provided the energy cost of replicating the whole thing does not become prohibitive, this does not seem to have a huge effect on the fitness of the organism; this is probably why there are amoeboid protozoans with genomes two orders of magnitude bigger than the human genome, though to my knowledge nobody has yet done bulk sequencing on those organisms. [ If you have a fairly simple replicating element that copies itself immense numbers of times, and you are working at a level wehre you sequence relatively short sections of the DNA and then try to assemble the bits of sequence later, trying to figure out what goes where when more than 90% of it is identical or almost-identical repeats of small pieces is nigh-impossible. ]

[nancylebov.livejournal.com]

http://genomicron.blogspot.com/2007/05/non-coding-dna-and-opossum-genome.html

One, very few people have actually argued that all non-coding DNA is 100% functionlesss "junk", and no one is surprised anymore when a regulatory or other function is observed for some non-coding DNA sequences. Moreover, transposable elements are more commonly labeled as "selfish DNA", and it has been noted in countless articles that they can and do take on functions at the organism level even if they begin as parasites at the genome level. Two, yet again we are talking about a small portion of the genome such that this should not be considered a demonstration that all non-coding DNA is functional.

Found that in the course of trying to find a different link about junk DNA, but looks like a writer worth reading.

http://www.getfinest.com/breaking_news/806.asp
Scientists at the Centre for Cellular and Molecular Biology (CCMB) here have demonstrated that junk DNA in human Y-chromosome control the function of a gene located in another chromosome.
Everything is more complicated than it looks. Biology is more like everything than anything else is.

http://www.lbl.gov/Science-Articles/Archive/mouse-dna-model.html

BERKELEY, CA — The already high value of the mouse as a model for studying the human genome has been raised even higher with the results of a new study by researchers with the U.S. Department of Energy's Lawrence Berkeley National Laboratory (Berkeley Lab) and the University of California at San Francisco (UCSF).

In a paper published in the April 7 issue of the journal Science, the researchers report that comparative analysis techniques used to identify DNA sequences coding for genes in mice and humans can also be used to identify sequences that regulate the "expression" or activation of genes.


"You could call this finding jewels in junk DNA," says Edward Rubin, a geneticist with Berkeley Lab's Life Sciences Division, and co-leader, along with Kelly Frazer, of the study. "By comparing human and mouse sequences we can identify those segments of the genome that contain information which instructs surrounding genes on when and where they are to be active. Identifying these sorts of regulatory sequences using classical biological approaches is labor intensive and difficult."

Evolutionary conservation of non-coding DNA sequences that play an important role in regulating gene expression is the key to the success of this study, just as it has been a key to identifying DNA sequences that code for genes across different species.

"If evolution conserved a sequence over the 70-90 million years since mice and humans diverged, it likely has a function," says Frazer. "Whether its function is to determine the structure of a protein coded for by a gene or to regulate gene expression, we should be able to identify these sequences through mouse to human sequence comparisons."

Unfortunately, I haven't been able to find a link for what I wanted--a news story I remember about junk DNA affecting the rate of production of proteins (or at least a protein?) which affecting how it folds.

[madfilkentist]

Human-written software that's gone through a lot of rewrites is generally full of code that no one understands and no longer serves any useful purpose. I wouldn't expect code generated by evolution to be any cleaner.

[dcseain.livejournal.com]

I agree.

[adrian-turtle.livejournal.com]

Where does the idea come from that "evolution doesn't make junk?" It seems silly to be that teleological about it. Evolution doesn't have an end in mind. It's not even like human-designed code, where humans know approximately what they're trying to do, and fumble around trying to optimize it. It's more like the proverbial large group of monkeys with typewriters, and the objective is not "optimal" but "survivable." Random changes to the code are very, very, often junk, when they aren't deadly.

[landley.livejournal.com]

Ooh, this takes me back. Specifically, to undergraduate molecular biology at Rutgers. :)

Lots of it codes for something called heat shock proteins".

When a cell gets overheated, several things happen. Random small holes get torn in everything (including the DNA) by brownian motion, in extreme cases proteins get denatured, and the control molecules get knocked off the DNA. The control molecules stop ribosomes from binding to the pribnow/tata sequences at the start of each gene and transcribing that gene into MRNA, from which protein gets made. (Lots of the dna transcription control mechanisms have to do with unlocking the appropriate control molecule bound to the gene's promoter.) So when a cell goes into heat shock, ribosomes start randomly transcribing all sorts of DNA for who knows what.

At this point, the cell's first priority (if it's to survive) is to repair its frayed DNA before the strands actually snap. So it wants DNA repair proteins (to go along the strand and pair up missing nucleotides), and lots of them, now. If ribosomes are transcribing any old thing, then the way to make sure you get DNA repair proteins out of the mess is to take the DNA repair protein gene and scatter about 300,000 copies of it all over the place, so that a significant chunk of the random junk that gets generated by the out-of-control machinery is DNA repair proteins. Then hopefully, by the time the cell gets itself back under control, the DNA has been salvaged and there's something to rebuild from.

So that's one source of random stuff that they thought was useless for the longest time until they figured out what it _did_...

Rob

[nancylebov.livejournal.com]

That's really cool. Is there anything known about how all the garbage generated by the recovery process gets cleaned up, or does it somewhat accumulate?

[landley.livejournal.com]

Google for it. Apparently, rather a lot of research has been done since my class in 1994. :)

http://www.tulane.edu/~biochem/med/hsp.htm

Rob