Register | Sign In


Understanding through Discussion


EvC Forum active members: 65 (9162 total)
6 online now:
Newest Member: popoi
Post Volume: Total: 915,815 Year: 3,072/9,624 Month: 917/1,588 Week: 100/223 Day: 11/17 Hour: 0/7


Thread  Details

Email This Thread
Newer Topic | Older Topic
  
Author Topic:   On the proportion of Nucleotides in the Genome and what it can tell us about Evolutio
slevesque
Member (Idle past 4640 days)
Posts: 1456
Joined: 05-14-2009


Message 46 of 61 (524585)
09-17-2009 3:51 PM
Reply to: Message 43 by Dr Jack
09-17-2009 4:42 AM


Well I would like to focus on eukaryotic cells, but I have difficulty finding data on it (Principally because I don't know where to look).
About the GC and AT triple and double bound, should it have an effect on DNA proportions of eukaryotic genomes ? (Considering what I said about inheritable mutations 'probably' happening during Meiosis ? (Which is just an intuitive guess on my part)

This message is a reply to:
 Message 43 by Dr Jack, posted 09-17-2009 4:42 AM Dr Jack has seen this message but not replied

Replies to this message:
 Message 47 by sfs, posted 09-17-2009 4:34 PM slevesque has not replied

  
sfs
Member (Idle past 2533 days)
Posts: 464
From: Cambridge, MA USA
Joined: 08-27-2003


Message 47 of 61 (524588)
09-17-2009 4:34 PM
Reply to: Message 46 by slevesque
09-17-2009 3:51 PM


Everything I have seen suggests that differences in GC content result from biases in mutational processes. These include biased gene conversion and biased repair mechanisms. The GC levels for eukaryotes can vary a lot: GC is ~43% for humans and ~20% for P. falciparum (malaria).

This message is a reply to:
 Message 46 by slevesque, posted 09-17-2009 3:51 PM slevesque has not replied

  
Wounded King
Member
Posts: 4149
From: Cincinnati, Ohio, USA
Joined: 04-09-2003


Message 48 of 61 (524590)
09-17-2009 4:55 PM
Reply to: Message 45 by slevesque
09-17-2009 3:49 PM


Actually, creationists are continuously pointing out to these advancements to show that 'Junk DNA' isn't junk after all. I would often argue that this strawman of modern biology is actually being done by the evolutionists
Do you think that these advancements arose from creationist research? What is the most recent paper you can find propounding Ohno's original formulation, I doubt you could find anything much more recent than the eighties. Certainly by the time I went to University in the 90s the roles of numerous non-coding regions were appreciated and taught in an evolutionary context, hardly maintaining a strawman of non-functional 'Junk' DNA.
'Junk' DNA was one of those frequent instances where a scientist made some predictions based on an incomplete understanding of the system. Even with the new functional transcribed non-coding elements and long distance regulatory elements that we are aware of there is still plenty of microsatellite DNA and highly repetitive regions to perform the sort of non-functional functions, like spacing between genes or surrounding centromeric regions, that Ohno suggested 'Junk' DNA had. The main thing Ohno had wrong was thinking that 'Junk DNA' was principally derived from 'failed' genes, there are certainly many pseudogenes which may be degenerate gene duplicates but the larger proportion of 'Junk' DNA originates from transposons. Even these transposon originated sequences serve some regulatory functions occasionally.
As you say Kimura and Haldane are another topic.
TTFN,
WK

This message is a reply to:
 Message 45 by slevesque, posted 09-17-2009 3:49 PM slevesque has not replied

  
Richard Townsend
Member (Idle past 4732 days)
Posts: 103
From: London, England
Joined: 07-16-2008


Message 49 of 61 (524604)
09-17-2009 7:05 PM
Reply to: Message 24 by Dr Adequate
09-16-2009 10:44 AM


Re: Statistics --- ur doin it rong
Dr A, can you modify your program to also calculate what the average excess of heads or tails is to verify the root n prediction?

This message is a reply to:
 Message 24 by Dr Adequate, posted 09-16-2009 10:44 AM Dr Adequate has replied

Replies to this message:
 Message 53 by Dr Adequate, posted 09-17-2009 11:20 PM Richard Townsend has not replied

  
Dr Adequate
Member (Idle past 284 days)
Posts: 16113
Joined: 07-20-2006


Message 50 of 61 (524627)
09-17-2009 9:57 PM
Reply to: Message 44 by slevesque
09-17-2009 3:40 PM


But will it realy make a difference in the long term ? I mean, if we take a genome and look at it mutate for a long time. Will not every nucleotide position, after several mutations, have a 50% chance of being AT or GC, and so when we consider the total ratio of the genome, we should expect to find approximately a 50/50 combination of AT and GC in it ?
Well, that depends on exactly what's going on. If we have a scenario as in "case 2" in my previous post, then no. In that case the AT content will have an equal chance of taking on any value. Note that this does agree with your intuition that "every nucleotide position, after several mutations, [will] have a 50% chance of being AT or GC" --- but only because 50% is the average of the numbers between 0% and 100%. 50% would therefore be the "expected value" only in the statistical sense of being the mean average outcome, not in the sense that we should particularly expect that the AT content should have that value rather than any other.

This message is a reply to:
 Message 44 by slevesque, posted 09-17-2009 3:40 PM slevesque has not replied

  
sfs
Member (Idle past 2533 days)
Posts: 464
From: Cambridge, MA USA
Joined: 08-27-2003


Message 51 of 61 (524631)
09-17-2009 10:32 PM
Reply to: Message 41 by Dr Adequate
09-16-2009 5:46 PM


quote:
Consider the following two cases.
Case #1 : We have a genome. We keep tossing a coin, and every time it comes up heads we add an AT base-pair, and every time it comes up tails we add a CG base-pair.
Now, this is not what you're trying to model. But it's what you're modeling. When you appeal to the Law Of Large Numbers in the way that you do, that's what you're modeling.
Case #2 : We have a genome of a fixed length. We keep tossing a coin, and every time it comes up heads we change a CG base-pair into an AT base-pair, and every time it comes up tails we change an AT base-pair into a CG base pair.
You probably shouldn't be modeling that, either. But that would be a "random walk" and wouldn't tend to anything.
You're correct: you shouldn't be modeling either case if you're interested in the mutation question. If the number of A/T pairs is currently Na and the number of G/C pairs is Ng, the probability that a new mutation will increase the AT content is 0.67 * Ng / (Na + Ng) and the probability that it will increase the GC content is 0.67 * Na / (Na + Ng). This process is indeed biased toward 50% GC.

This message is a reply to:
 Message 41 by Dr Adequate, posted 09-16-2009 5:46 PM Dr Adequate has replied

Replies to this message:
 Message 52 by Dr Adequate, posted 09-17-2009 10:51 PM sfs has replied

  
Dr Adequate
Member (Idle past 284 days)
Posts: 16113
Joined: 07-20-2006


Message 52 of 61 (524637)
09-17-2009 10:51 PM
Reply to: Message 51 by sfs
09-17-2009 10:32 PM


Well, that depends again on what exactly is going on. What you're modeling there is where the mechanism for mutation is "Pick a base at random, then change it with equal probability to any base it isn't". This is somewhat simplistic.
I have done a model of that case, and you are right: the probability distribution does indeed spike around 50% AT. I also find that biases in the probability of mutations and their repair shift this spike with relative ease: if, for example, a change from AT -> CG is twice as likely to be repaired as one from CG -> AT, then this shifts the spike to more like 75%.

This message is a reply to:
 Message 51 by sfs, posted 09-17-2009 10:32 PM sfs has replied

Replies to this message:
 Message 54 by Dr Jack, posted 09-18-2009 4:40 AM Dr Adequate has not replied
 Message 58 by sfs, posted 09-18-2009 7:40 AM Dr Adequate has not replied

  
Dr Adequate
Member (Idle past 284 days)
Posts: 16113
Joined: 07-20-2006


Message 53 of 61 (524641)
09-17-2009 11:20 PM
Reply to: Message 49 by Richard Townsend
09-17-2009 7:05 PM


Re: Statistics --- ur doin it rong
It seems that it would be more accurate to say that it's of the order root n (using "big O" notation).
Here's a graph of values for n between 1 and 10000:
The blue line is the square root of n, the red line is the average of |h(n) - t(n)| over ten thousand trials.

This message is a reply to:
 Message 49 by Richard Townsend, posted 09-17-2009 7:05 PM Richard Townsend has not replied

Replies to this message:
 Message 57 by Peepul, posted 09-18-2009 7:01 AM Dr Adequate has not replied

  
Dr Jack
Member
Posts: 3514
From: Immigrant in the land of Deutsch
Joined: 07-14-2003
Member Rating: 8.7


Message 54 of 61 (524681)
09-18-2009 4:40 AM
Reply to: Message 52 by Dr Adequate
09-17-2009 10:51 PM


Because of deamination (in which cytosine loses an amine group, and is converted to uracil which then binds to thymine rather than guanine during DNA replication) it sees quite probable that GC -> AT point mutations are more likely than AT -> GC point mutations.

This message is a reply to:
 Message 52 by Dr Adequate, posted 09-17-2009 10:51 PM Dr Adequate has not replied

Replies to this message:
 Message 55 by Wounded King, posted 09-18-2009 5:03 AM Dr Jack has replied

  
Wounded King
Member
Posts: 4149
From: Cincinnati, Ohio, USA
Joined: 04-09-2003


Message 55 of 61 (524685)
09-18-2009 5:03 AM
Reply to: Message 54 by Dr Jack
09-18-2009 4:40 AM


Nit-picking
That needs to be 5-methyl cytosine for it to really work. There are enzymes which specifically excise and replace uracil bases in the DNA. The repair system may still allow a few through though. 5-Methyl cytosine changes to thymine which can't be distinguished.
Also uracil binds to adenine not thymine.
TTFN,
WK

This message is a reply to:
 Message 54 by Dr Jack, posted 09-18-2009 4:40 AM Dr Jack has replied

Replies to this message:
 Message 56 by Dr Jack, posted 09-18-2009 5:11 AM Wounded King has not replied

  
Dr Jack
Member
Posts: 3514
From: Immigrant in the land of Deutsch
Joined: 07-14-2003
Member Rating: 8.7


Message 56 of 61 (524686)
09-18-2009 5:11 AM
Reply to: Message 55 by Wounded King
09-18-2009 5:03 AM


Re: Nit-picking
Yes, sorry, binds to uracil, thus replacing the cytosine with thymine *sigh*
That needs to be 5-methyl cytosine for it to really work. There are enzymes which specifically excise and replace uracil bases in the DNA. The repair system may still allow a few through though.
And, yes, most such changes are caught, but not all. Which should provide a bias to the probability to GC->AT vs. AT->GC changes.

This message is a reply to:
 Message 55 by Wounded King, posted 09-18-2009 5:03 AM Wounded King has not replied

  
Peepul
Member (Idle past 5017 days)
Posts: 206
Joined: 03-13-2009


Message 57 of 61 (524699)
09-18-2009 7:01 AM
Reply to: Message 53 by Dr Adequate
09-17-2009 11:20 PM


Re: Statistics --- ur doin it rong
Yes that's exactly what it looks like! Cool programming!
Edited by Peepul, : No reason given.

This message is a reply to:
 Message 53 by Dr Adequate, posted 09-17-2009 11:20 PM Dr Adequate has not replied

  
sfs
Member (Idle past 2533 days)
Posts: 464
From: Cambridge, MA USA
Joined: 08-27-2003


Message 58 of 61 (524703)
09-18-2009 7:40 AM
Reply to: Message 52 by Dr Adequate
09-17-2009 10:51 PM


quote:
Well, that depends again on what exactly is going on. What you're modeling there is where the mechanism for mutation is "Pick a base at random, then change it with equal probability to any base it isn't". This is somewhat simplistic.
Oh, absolutely -- it's very simplistic. For a more realistic model you would substitute a transition probability matrix, various of which are floating around the literature. But it is at least a simplistic model of mutational substitutions, which is not true of either of the other models listed.

This message is a reply to:
 Message 52 by Dr Adequate, posted 09-17-2009 10:51 PM Dr Adequate has not replied

  
slevesque
Member (Idle past 4640 days)
Posts: 1456
Joined: 05-14-2009


Message 59 of 61 (524840)
09-19-2009 2:37 PM


So to recap what has been said, can someone answer this question;
What ratio GC/AT should we expect from a mutating genome ?
If we can answer this question, we can add the effects NS later on.
PS. A totally acurate answer is probably to much to ask, an approximte and qualitative answer should do.
EDIT: sfs, are you the same person as sfs1 on Tweb ?
Edited by slevesque, : No reason given.

Replies to this message:
 Message 60 by Dr Adequate, posted 09-19-2009 4:35 PM slevesque has not replied
 Message 61 by sfs, posted 09-20-2009 4:55 PM slevesque has not replied

  
Dr Adequate
Member (Idle past 284 days)
Posts: 16113
Joined: 07-20-2006


Message 60 of 61 (524852)
09-19-2009 4:35 PM
Reply to: Message 59 by slevesque
09-19-2009 2:37 PM


So to recap what has been said, can someone answer this question;
What ratio GC/AT should we expect from a mutating genome ?
The only answer anyone can give you would be: similar to that of a closely related species. For example, the figure for humans should be close to that for chimpanzees. But there is seemingly no a priori method by which one can readily predict the AT content of a species.

This message is a reply to:
 Message 59 by slevesque, posted 09-19-2009 2:37 PM slevesque has not replied

  
Newer Topic | Older Topic
Jump to:


Copyright 2001-2023 by EvC Forum, All Rights Reserved

™ Version 4.2
Innovative software from Qwixotic © 2024