Re: Any real evidence for evolution, point on point.
quote:
Well I guess you better notify Michael W. Nachman, and Susan L. Crowell that the units do not match in the equation they used for this paper. By the way do you know the units of (k) and (u)?
Go back and read the post immediately before yours. It gives the units, and also explains why Nachman and Crowell's equation is just fine for single-base substitutions. Think about it until you understand it.
quote:
k = is mutations (%)
Which is it? The number of mutations or the percentage difference? If you have two identical genomes, except that one of them has had an insertion mutation of 30 million base pairs, 1 out of 100 bases is different, but only 1 out of 3,000,000,000 sites has mutated. The mutation rate counts this as 1 event, while the percentage difference counts it as 30,000,000 differences. Counting the number of mutations is never going to get you to the percentage difference unless you know how big the mutations are.
quote:
Are you claiming interbreeding populations (by the way the populations at this point were relatively large) did not have a female female linage? That is what would have to happen again and again.
The population in question -- the people migrating out of Africa -- was small, and probably only a handful of matings between anatomically modern humans took place. Only about a third of the Neandertal genome is preserved in the descendants of those matings; there's no reason that mitochondrial DNA should be part of it.
quote:
You accuse me of having improbable mechanisms? That is total nonsense.
Show your work: given a plausible demographic model of the Out of Africa migration, and estimates for the amount of Neandertal admixture, calculate the probability that Neandertal mtDNA would have survived.
Re: Any real evidence for evolution, point on point.
quote:Units are mutations and represent variance is percentage.
No, mutations represent mutations. How much divergence they represent depends on big they are. This is simply a fact. There's no opinion here: what you've been saying is wrong.
quote:
OK let us count every site difference between the two genomes and see what percentage of variance comes up. The number is bp adjusted via alignment tool.
Fine. To go along with that, we'll also need the mutation rate in base pairs per generation. Do you know what it is for indels? It's not the rate you've been quoting -- that's the rate of mutations per generation. If you don't know it, you can't do the calculation.
quote:
Again bp is not completely counted, it is defiantly the adjusted percentage difference between the human chimp genome. Look, if you did a bruit force comparison between base pairs, human against chimp, the similarity of base pairs would be in the 65% range. You do not want to go there. So saying that all base pairs are accounted for in variance is just not true.
I used to write algorithms for variance (not for biology). The alignment tools used in these comparisons adjust for distances and gaps and bp, in the genomes.
They don't adjust anything. Alignment tools align the parts that align and show gaps where they don't.
Re: Any real evidence for evolution, point on point.
quote:
(indels) are also quantifiable under neutral model calculations.
Of course they are. But since we don't have a good direct estimate of their mutation rate, you can't independently check to see whether human/chimpanzee divergence in indels fits with the mutation rate. There's nothing to compare it to.
quote:
The majority of all papers dealing with indel variation, directly or indirectly, note that indels must be included in percentage similarity.
The vast majority of papers dealing with indels having to do with similarity. Indeed, the overall percentage similarity is of very little scientific interest.
quote:
The paper in question, discounted indels because the paradigm had not changed at that time.
What paper discounted indels? And what paradigm changed?
quote:
Indels must be included in the divergence calculation or you must throw out the entire premise of that calculation.
That depends on what you're calculating and why. So far, in this thread, you haven't been calculating anything connected to the real world.
Re: Any real evidence for evolution, point on point.
quote:
I don't see why they wouldn't be included. The problem is that rare but large indels can really skew the data. For the human and chimp divergence, the rare but large indels account for a hefty portion of the differences while small but plentiful indels account for less of the differences.
That's one issue. Another is that it's much harder to accurately detect indels, especially ones that are more than a few base pairs. The direct estimate of mutation rates in humans that I'm most familiar with (this one) only looked at single base substitutions. They're just a lot easier to deal with.
Re: Any real evidence for evolution, point on point.
quote:
Is this due to the difficulty of aligning gappy sequence and distinguishing indels from DNA that wasn't sequenced? Could these problems by solved by using slower, but more accurate sequencing methods on BAC clones?
I'm not a sequencer, so I don't know all of the issues. With diploid sequence, you'll have some reads with and some without the indel. If you're doing lowish coverage, short-read sequencing, there may not be very many reads of both type, and the indel is going to be near the end of some of them, making it hard to distinguish indels from SNPs. BACs would definitely be easier, since they're haploid. Of course, you'd need to sequence multiple BACs from the same chromosome segment, since you'd need to have BACs with both versions of the sequence.
Re: Any real evidence for evolution, point on point.
quote:
I am sure this is accurate for what genes this paper looked at. These numbers do vary from paper to paper according to the focus of the researchers.
Which part of "largely complete" did you not understand? The paper looked at (nearly) the entire genome.
quote:
I agree for the gene segments analyzed.
Which was nearly all of them.
quote:
I would say you can not get to 5% from 1.33% in these results. Like I say, different findings for different genes investigated.
Of course you can get to 5% -- provided you account for the size of the indels. But you've only been told that 10 or 20 times now, so I'm sure you're not going to understand it yet. In fact, this paper does say the difference is 5%. (Well, actually it says something more accurate than that, but let's try to stick to the simplest story here.) And no, it's not because of the genes that were investigated. (In fact, genes constitute only a tiny fraction of the genomes being compared.)
quote:
Here is a shameless repost of the calculation I did with only using indel variables from the proceeding paper and reduced mutation rate for indels.
Incorrect calculations don't become correct by repeating them.
quote:
As I have stated over and over indel and substitution rates are addable. They both reflect autosomal sequence divergence for the (k) in the calculation in question.
And as people who know far more about the subject than you do have replied over and over, you're wrong.
I do wonder about the source of this idea that anybody's interpretation of a scientific paper is a valid as anybody else's. I wonder if some kind of misbegotten offspring of the protestant doctrine that everyone should be able to read and interpret scripture for themselves. I dunno. In any case, you're claiming to understand these papers better than their authors, which is kind of silly.
Re: Any real evidence for evolution, point on point.
quote:
Believe it or not sfs and I have hashed this out over a year ago. I never claimed 0% homology. By the same token sfs could not claim (100%) homology.
If by "hashed out", you mean that you made rambling, inaccurate claims, I corrected you, and you ignored me, then yes, we hashed it out.
Re: Any real evidence for evolution, point on point.
quote:
I remember I maintained a point on several issues.
If I remember correctly, your little program you wrote was using asexual reproduction rates contrary to our sexual reproduction discussion, your similarity in protein coding segments was not using the poisson distribution like you should have. etc.
The past is the past, I always say....Ohh... by the way I did prevail in the discussion on alignment.
I don't know how many encounters we've had. The one I found is the one where I wrote the following: "What you've written is complete gibberish." It was. Everything else you wrote in that thread was also either wrong or completely meaningless.
Of course you think you prevailed. That's because, as is obvious from this thread, that you understand so little of what you're talking about that you can have no way of telling when you're wrong.
Re: Any real evidence for evolution, point on point.
quote:So does the chimp genome paper if you are counting number of bases instead of number of mutations. The chimp genome paper has 5 million indels covering 90 million bases compared to 35 million substitution mutations.
Amounting to 1.5% of the human sequence being different from the chimp sequence and another 1.5% of the chimp being different from the human. Added to the 1.3% from substitutions, and you get ~4.3% divergence(*). As the paper says, "Insertion and deletion (indel) events are fewer in number than single-nucleotide substitutions, but result in 1.5% of the euchromatic sequence in each species being lineage-specific."
All the studies give similar answers.
(*)No, you really shouldn't measure divergence this way.
quote:
If they are added correctly, then I can use them as autosomal divergence, right.
Quite right. Unfortunately, you are not able to add them correctly. You've been trying for years, and you still can't add them correctly. You've been given trivial examples, and you can't parse them correctly. Perhaps you should find something else to do with your time.
Re: Any real evidence for evolution, point on point.
quote:
There are 70 new substitutions each generation. If the indel rate is 1/7th of that as sfs indicated, then that is 10 indels.
Note that all I did was take the 5 million indels from the chimpanzee genome paper and divide by the 35 million substitutions. (Also, I suspect the reason he's tried to denigrate the chimp genome paper and suggest it's been superseded is that he knows I'm one of the authors.)
quote:
Here is the simple version of divergence time used in calculating difference between species Without Ne. A simple approximation. I like simple. d = 2*r*t or t = .5(d/r) where d = DNA differences between two individuals r = the measured mutation rate in the species or lineage t = time of origin derived from each origins model (in generations)
Simple is good. Let's use a cartoon organism. It has a mutation rate of 1x10^-9/bp/gen and a genome of 1 billion base pairs, so on average there is 1 mutation per generation. Half of its mutations are single-base substitutions, and half are large indels. Let's look at a copy of the genome from each branch just one generation after they split. In that 1 generation, one copy acquired one single-base substitution. The other copy acquired an insertion of 10 million base pairs. As a result, the genomes now differ by 10,000,001 base pairs, or just over 1%.
Using your formula, t = 0.5*(.01/1e-9) = 5,000,000 generations. That's wrong by a factor of 5 million. What do you think might be wrong with your formula?
™ Version 4.2
(1)