ALL BLOG POSTS AND COMMENTS COPYRIGHT (C) 2003-2019 VOX DAY. ALL RIGHTS RESERVED. REPRODUCTION WITHOUT WRITTEN PERMISSION IS EXPRESSLY PROHIBITED.

Monday, March 18, 2019

Evolution is out of time

I discussed this a bit on last night's Darkstream, but because my grasp of the technicalities of how genetics work is close to nonexistent, I didn't even try to delve into the details. It's much better to simply read the linked articles; I leave it to those more versed in the subject to determine how valid the reports of the massive gap between the oft-reported 98 percent estimated similarity between the chimp and human genomes and what genetic scientists are actually seeing now as their ability to analyze the various genomes improves.

The first exhibit is an interview with a creationist geneticist, which will no doubt be improperly dismissed by scientistry fetishists with an appeal to the genetic fallacy.
Dr. Tomkins: My motivation started when I arrived here and was given the task of researching the human-chimpanzee similarity issue because people ask about this in churches. They hear the claim that humans and chimps are 98 to 99% similar. People want to know if that’s true. Before working here, I’d not investigated that issue. I ran a genome center for over five years and investigated various plants and animals but never the human-chimpanzee comparison. I went into it with an open mind and began reading all the literature on the subject—this started about eight years ago. I looked at the top six scientific publications that proposed a 98 to 99% DNA similarity between modern humans and modern chimpanzees.

Brian: A 98 to 99% genetic similarity between modern humans and modern chimps—why is that important?

Dr. Tomkins: It’s very important to theoretical evolutionists. The 98 to 99% claim is a theory—it’s speculative. They need a similarity that close to have humans and chimps evolve in the alleged three- to six-million-year timespan from a supposed human-chimpanzee common ancestor. Their statistical models need that 98 to 99% similarity.

Brian: What did you find in the literature?

Dr. Tomkins: The first thing I noticed when I began reading these articles was that researchers were throwing out a lot of data. They were cherry-picking the areas of DNA between humans and chimps that were highly similar and throwing out areas, including areas that would not line up properly. Areas that don’t line up are dissimilar. When I researched the data, I was coming up with DNA similarities between 81 to 86% when I included the dissimilar data. I published a paper on this.1 This is way outside the realm of theoretical evolution.

Brian: What should the evolutionary community say about this?

Dr. Tomkins: They have reacted to a lot of my research since that first paper. There’s a lot of DNA sequence data that is publicly available in databases. I began working with the data myself, and over a number of years I refined my techniques. I used an algorithm developed by evolutionists that turned out to be a bad algorithm—so there’s been a lot of trial and error. But I finally got to the point where I published a paper in 2016.2 It was the most comprehensive study I’ve done yet, and I looked at all 101 data sets that went into originally building the chimpanzee genome.

I sampled 25,000 sequences at random from each of the data sets and then began analyzing and comparing them to human. Over half of the data sets were extremely similar to human, and the other half were extremely dissimilar to human. It appeared the initial chimpanzee genome was contaminated with human DNA, which is a huge problem in genomics.

There’s a number of studies by secular researchers showing that many public DNA databases, from bacteria to fish, have significant levels of human contamination. Human DNA literally gets into the samples. Contamination is a major issue. Human DNA comes from researchers’ fingers, coughing, sneezing, etc., and it gets into the samples. Now researchers are taking greater steps to alleviate that problem. This was especially prevalent back in the earliest phases of genome projects, when the chimpanzee was sequenced.

Brian: Wouldn’t some of the human DNA that made it into the raw data affect the results of any comparison analyses?

Dr. Tomkins: It has a huge effect because the chimpanzee genome is stitched together using the human genome as a scaffold. It’s like a puzzle—researchers used the human DNA “picture on the box” to assemble the chimp genome. The chimp DNA sequences used were all about 750 bases long. Not only was the chimp genome built using the human genome as a guide, it also has human DNA contamination in it, so it showed a lot of similarity from the contamination.

Brian: Even with those factors in place that skewed the data to a more human genome, is it closer to the 98% or the 86% maximum you observed?

Dr. Tomkins: It’s difficult to determine because it is a flawed product. I based my research on human-chimp similarity on the half of the data sets that appear to have much less human DNA. Based on my work, I’m seeing not more than an 85% DNA similarity of chimpanzee to human, and that’s a maximum. It’s probably less than that.
The second exhibit is even more interesting, because an evolutionary biologist who is the Professor of Evolutionary Genomics at the University of London has been seeing much the same thing in his review of the various chimp-human genomic studies:
When assessing the total similarity of the human genome to the chimp genome, we also need to bear in mind that roughly 5% of the human genome has not been fully assembled yet, so the best we can do for that 5% is predict how similar it will be to the chimpanzee genome. We do not yet know for sure. The chimpanzee genome assembly is less well assembled, so in future we may assemble parts of the chimpanzee genome that are similar to the human genome – this is another source of uncertainty to keep in mind.

To come up with the most accurate current assessment that I could of the similarity of the human and chimpanzee genome, I downloaded from the UCSC genomics website the latest alignments (made using the LASTZ software) between the human and chimpanzee genome assemblies, hg38 and pantro6. See discussion post #35 for details. This gave the following for the human genome:

4.06% had no alignment to the chimp assembly
5.18% was in CNVs relative to chimp
1.12% differed due to SNPs in the one-to-one best aligned regions
0.28% differed due to indels within the one-to-one best aligned regions

The percentage of nucleotides in the human genome that had one-to-one exact matches in the chimpanzee genome was 84.38%

In order to assess how improvements in genome assemblies can change these figures, I did the same analyses on the alignment of the older PanTro4 assembly against Hg38 (see discussion post #40). The Pantro4 assembly was based on a much smaller amount of sequencing than the Pantro6 assembly (see discussion post #39). In this Pantro4 alignment:

6.29% had no alignment to the chimp assembly
5.01% was in CNVs relative to chimp
1.11% differed due to SNPs in the one-to-one best aligned regions
0.28% differed due to indels within the one-to-one best aligned regions

The percentage of nucleotides in the human genome that had one-to-one exact matches in the chimpanzee genome was 82.34%.

Thus the large improvement in the chimpanzee genome assembly between PanTro4 and PanTro6 has led to an increase in CNVs detected, and a decrease in the non-aligning regions. It has only increased the one-to-one exact matches from 82.34% to 84.38% even though the chimpanzee genome assembly is at least 8% more complete (I think) in PanTro6.
I have already shown that it is highly improbable that the speed of mutational fixation is sufficient to account for the estimated 98 percent similar relationship which needs to account for 30 million fixed mutations since the Last Chimp-Human Common Ancestor, so this massive increase in the observed difference between the two genomes, which is presently calculated to end up somewhere between 84.38 to 93.43 percent, is enough to not only drive the final nail in the Neo-Darwinian coffin, but wrap it in iron bands, encase it in concrete, and drop it into the Marianas Trench.

Because what we're seeing here is inept statistical wizardry that involves everything from contaminated evidence to cherry-picked data and the ridiculous assumption that literally ALL of the remaining unknown areas will, in the future, be found to perfectly align with orthodox Neo-Darwinian theory. And the priests of TENS are still desperately clinging to that improbable assumption even though it was only proved to be correct for 25.5 percent of the area that was filled in over the course of the seven years that passed between the publications of PanTro4 and PanTro6.

Of course, no one here will be even remotely surprised to observe that Prof. Buggs's original 2008 prediction was too low because, in the 2005 paper upon which he relied, the biologists got the math wrong.

Labels: ,

78 Comments:

Blogger maniacprovost March 18, 2019 7:29 AM  

We know so little about the genome. If they reliably isolated tgt non-functional DNA only, the comparison of those segments should give you a better clock.

Blogger D E K March 18, 2019 7:47 AM  

Mind blowing. As someone who never was so much interested as to take a closer look it is almost shocking. I guess all the supposed similarities with pigs and flies are going to be corrected down in the near future. Sneezing into probes and trying to build something with the human genome as a blueprint...who would have thought.
It will end up with "long live the engineers and to hell with the scientists and science"

Blogger wreckage March 18, 2019 8:04 AM  

So, bets that they end up scrapping the human-chimp common ancestor? It'll be funny if the new, properly accepted Science That Disproves God, like the Big Bang, actually turns out to have been prompted by Christian rejection of the accepted model. Again.

Blogger fiendeJ March 18, 2019 8:36 AM  

If the biologists get the math right, is it still biology?

Blogger Gregory the Tall March 18, 2019 8:49 AM  

JF: 99 percent, 85 percent what does it matter, we will just need to add a few other multiverses with warp gene transfer portals between all of them and it will be plausible.

Blogger Gettimothy March 18, 2019 9:05 AM  

@4 funny! spewed my coffee. no dna samples cotaminated.


Blogger Dad29 March 18, 2019 9:10 AM  

...Because what we're seeing here is inept statistical wizardry that involves everything from contaminated evidence to cherry-picked data and the ridiculous assumption that literally ALL of the remaining unknown areas will, in the future, be found to perfectly align with orthodox... anthropocentric global warming theory.

Wonder of wonders!

Blogger Silent Draco March 18, 2019 9:11 AM  

fiendeJ, bad dog! Extra biscuit!

Fortunately I learned to put the coffee down first, before reading comments. Scalded nose doesn't code well on the ER.

Blogger Gregory the Tall March 18, 2019 9:20 AM  

@6 Are you so sure it was not contaminated? Get a gene profile of that coffee, it will authoratively prove that an important not so distant ancestor of us humans is Coffea arabica, hence the scientific term "Arabica advantage" which defines the Darwinian fight for survival between Muslim and Christian societies.

Blogger JD Curtis March 18, 2019 9:22 AM  

David Berlinski approached the 'time' angle in another way.
He took a long look at what it would take in terms of profound morphological changes to the land dwelling creature Pakicetus to a become a full time, ocean dwelling whale. Here's the video if anyone is interested. If it doesn't come up in your area, you can search for it on YouTube by typing [David Berlinski what does it take for change clip 5] in the search field.

Blogger FUBARwest March 18, 2019 9:26 AM  

It's fascinating and a bit disheartening to find out so much of modernity is built on too of lies. Interesting times we live in now that a lot of those lies are being shown for what they are all at the same time. It is going to take decades, or atleast until after the collapse, for this particular brainwashing to be fixed.

The belief in science! is completely faith based for most people. I had a South African tell me he believes in TENS but doesn't think different humans are different. No answer to how groups of humans living different environments for centuries wouldn't adapt differently. Would have thought someone from SA would be a little more red pilled on the situation.


Completely OT: Vox do you know the title of the music that accidentally went off during last nights darkstream? It sounded like a good song.

Blogger Gregory the Tall March 18, 2019 9:27 AM  

@10 Berlinski is very good, I think. I read his book on the probability of the existence of God, and like Vox he destroys all the "irrational atheists" without mercy from a slightly different angle.

Blogger JD Curtis March 18, 2019 9:30 AM  

Right. And the guy's an agnostic, at best

Blogger The Cooler March 18, 2019 9:34 AM  

I share 50% of my genes with a banana. I am an American, therefore.

Blogger szook March 18, 2019 9:52 AM  

Out of Africa is an always was a myth built on the genome studies. All archaeology leads to the ME and the fertile crescent. All the assertions to ancient hominid decent from Africa was only ever possible because researchers admixed the genomes. Funny it's one thing to know they thought it was appropriate to do so with the meta-data.....it's quite another thing to learn they were doing it all along due to incompetent lab technique.

Blogger Gettimothy March 18, 2019 9:55 AM  

@14 I feeeeel like a banana.

Blogger grendel March 18, 2019 9:57 AM  

Check out "Replacing Darwin" by geneticist Nathaniel Jeanson.

Blogger Gregory the Tall March 18, 2019 10:10 AM  

@16 And there we have the scientifically sound answer to why chimps are drawn to bananas.

Blogger Daniel Babylon March 18, 2019 10:13 AM  

Can anyone calculate how many fixed mutations per generation would be required to account for a 15% difference?

Blogger Daniel Babylon March 18, 2019 10:14 AM  

If the original was 44 fixed mutations at 2%, then I get 330 mutations at 15%. Is that correct?

Blogger Ransom Smith March 18, 2019 10:24 AM  

it will authoratively prove that an important not so distant ancestor of us humans is Coffea arabica
Are you calling Vox a cannibal?

Blogger Blume March 18, 2019 10:40 AM  

That's what I get Daniel

Blogger One Deplorable DT March 18, 2019 10:44 AM  

Next time some !SCIENCE! nut tells me about the similarity of human and chimp DNA I'm not even going to bother with the evidence. I'm just going to pull out my iPhone and play this song.

Blogger Gregory the Tall March 18, 2019 10:45 AM  

@21 Is there something in the Bible suggesting one should rather not be one?

Blogger Francis Parker Yockey March 18, 2019 10:51 AM  

I have already shown that it is highly improbable that the speed of mutational fixation is sufficient to account for the estimated 98 percent similar relationship which needs to account for 30 million fixed mutations since the Last Chimp-Human Common Ancestor

Not really. See my comments #208, 209, & 211 under your previous post. To recap, here are some of the problems with the original argument:

You generated this stat --
Generations per fixed mutation: 1600 -- from the long-running Plenski E. coli experiment
without really understanding where it comes from, or what it means. A few issues with it:

1. You're attempting to conflate two separate, independent steps -- the generation of variation, and selection -- into one.

2. You're taking the term "generation" and using it as if it measures the same thing with respect to mutation rates in E. coli that it does in humans. Everything else held equal, mutation rates vary with the number of base pairs replicated, and the "generations" of genome replication, not whole organism replication.

There's big difference in the precise meaning of "generation" when applied to E. coli vs. humans. E. coli is single-celled. For single-celled organisms -- whether prokaryotic or eukaryotic -- a generation is a single cell division. A single genome replication. For humans, a generation represents many replications of the entire genome. There are about 20-30 genome replications between the fertilized egg and the female gametes, and about ten times that for males -- with large variation depending on age. That's roughly two orders of magnitude difference right there.

3. The E. Coli genome is roughly 1000x the size of the human genome -- that's another 3 orders of magnitude. 5 total.

4. The human mutation rate has been measured and turns out to be about 70 mutations/ individual/ generation, or 1-1.2 x 10^-8 per base pair per generation. Measured by whole-genome sequencing of successive generations (2010). This is about 2 orders of magnitude greater than measured mutation rates per base pair per generation for E. coli -- 8.9 × 10^−11 per base-pair per generation -- for the reasons discussed above.

Summary:

Bionumbers mutation rate per human generation

Full text

Blogger One Deplorable DT March 18, 2019 10:54 AM  

@11 - It's fascinating and a bit disheartening to find out so much of modernity is built on too of lies.

When has it been otherwise? Absent any real, physical constraints people are free to make up and believe whatever stories and narratives they desire. The scientific method is a useful but abstract tool. It only restrains and guides those with the integrity to follow it to the letter no matter the results.

That's why engineering is so much more reliable than !SCIENCE! In engineering it's not enough to say your plane design will fly or your bridge design will stand. Your plane actually has to fly and your bridge actually has to stand.

In what other fields is this true? Do politicians actually have to solve problems? Do doctors actually have to make you well? Do feminists actually have to improve the lives of women? Do economists actually have to improve the economy?

And what are the consequences if they don't?

We live in a world of lies because it's so easy, and profitable, to lie in so many fields of human endeavor.

Blogger Francis Parker Yockey March 18, 2019 10:59 AM  

In Vox's previous post, he cited the source paper for his stat on E. coli mutations in this manner:

Source: Sequencing of 19 whole genomes detected 25 mutations that were fixed in the 40,000 generations of the experiment.
NATURE, 2009


This doesn't include the title of the paper, but"Nature 2009" appears to be a reference to J. E. Barrick et al. Nature, 461:1243.
"Genome evolution and adaptation in a long-term experiment with E. Coli," a paper that came out of the long-term Lenski study. No one else has 40K generations (>60K now).

Genome evolution and Adaptation in a long-term experiment
[Paywalled except abstract]. That was the only paper the Lenski group published in Nature that year.

But the above quote appears to refer to a different (2011) paper from the same group:

2011 Lenski paper full text

Summary

Vox's phrasing is very similar to this quote from a summary that incorporates references to both the 2009 and 2011 papers -- in this instance, it's referring to the 2011 one:

Sequencing of 19 whole genomes detected 25 synonymous mutations (indicating neutral rather than selective changes) that got fixed in the 40,000 generations of the experiment.

But he seems to have omitted some critical words --bolded above -- from his characterization of the study's findings. The paper is looking at new synonymous mutations that progressed to fixation, not total mutations.

Synonymous mutation

Why does this matter?
The genetic code is highly redundant/ degenerate -- there are multiple nucleotide sequences that code for each amino acid. This paper is deliberately looking at new mutations that do not change the resulting protein, thus are not subject to selection -- yeah, some synonymous substitutions can have effects on transcription, or on RNA splicing if they're in the wrong location, but it's a reasonable approximation.

The intent of this part of the study is not to measure evolution rates, but to exclude evolutionary changes/ mutations that are actively selected for or against, in order to measure the underlying spontaneous mutation rate by measuring the appearance of new silent mutations in the population.
[Turned out to be about 8.9 × 10^−11 per base-pair per generation, which is consistent with other studies using different methodology]

The total number of new mutations that achieved fixation in the first 40,000 generations was 255. Pic related -- figure from yet another [2009] paper from the same group:
255 mutations achieved fixation

Note also that:
1. The E. coli strain chosen for this experiment has no plasmids or prophages -- strictly asexual reproduction, with no mechanism for horizontal transfer.

2. One of the mutations that achieved fixation in about half of the populations in the experiment was -- one that increased mutation rates roughly two orders of magnitude higher than the ancestral rate.

Blogger Blume March 18, 2019 11:02 AM  

Yancy I think your defense of evolution makes it even more unlikely that natural selection actually works as describe.

If nature selects me but my offspring are 1 out of 30 possible mutations of me then nature really hasn't selected anything.

Blogger Jeff Burton March 18, 2019 11:15 AM  

one that increased mutation rates roughly two orders of magnitude higher than the ancestral rate.

I suppose you see that as a good thing? It was achieved by damaging the bacteria's DNA repair mechanism by inserting a single erroneous base pair. Check out the abstract from Lenski's research on that: https://www.ncbi.nlm.nih.gov/pubmed/9192894.

Our results corroborate computer simulations of mutator evolution in adapting clonal populations, and may help to explain observations that associate high mutation rates with emerging pathogens and with certain cancers..

Doesn't sound very promising as an evolution accelerator, does it?

Blogger Blume March 18, 2019 11:20 AM  

I still think your defense is getting weaker. Your 20-30 replications between organisms would be of the synonymous type since mutations within the organism can't be selected for or against by either sexual or natural selection. Thus proving it is the correct rate for vox to have chosen. Plus if you think about it the synonymous mutation rate is the average you want since you can't actually know the plus or minus of historical selection.

Blogger Gen. Kong March 18, 2019 11:56 AM  

fiendeJ wrote:

If the biologists get the math right, is it still biology?

Only if they're using feminist or tranny math.

Blogger pyrrhus March 18, 2019 12:09 PM  

The time factor has never worked...Since less than 1/10*4 mutations is beneficial, and most mutations are harmful, it takes a staggering amount of time just to produce a few mutations useful enough to be selected for, and another staggering amount of time to have it fixed in a population...That's why the introversion of Neanderthal and Denisovan genes was so critical to the development of Homo Sapiens, it was a short cut.

Blogger rcocean March 18, 2019 12:11 PM  

Won't the TENS Supporters just expand the timeline again? Already its gone from a million years to a Billion Years. Now it will go to a Trillion years.

Blogger Gregory the Tall March 18, 2019 12:20 PM  

@32 and 33 evolutionists apparently adhere to the notion that time is an illusion anyway, therefore voluntarily shrinkable or expandable according to the scientists whims.

Blogger Fargoth March 18, 2019 12:21 PM  

We are still being taught the 98% figure in upper-level university genetics courses. Thanks Vox, I would have had no idea the sequence identity claim was this fraudulent but we shouldn't be surprised by this point.

Darwin's rotting carcass is swinging from the rafters....

Blogger VD March 18, 2019 12:23 PM  

Not really. See my comments #208, 209, & 211 under your previous post. To recap, here are some of the problems with the original argument.

No. You're completely incorrect. You didn't even touch upon the original argument. What you're doing is confusing the walk through someone's thought processes with the actual argument to which it led. Nothing you do to correct whatever mistakes I made with regards to the details of various e. coli studies has anything whatsoever to do with a) the number of fixed mutations necessary to transition from LCHCA to human or b) the amount of time involved.

So, how many mutations do you assert can be fixed per human generation?

I don't think you have grasped how completely this destroys evolution by natural selection. It's done. It's gone. It's finished. Because you're trapped in an unavoidable dichotomy. Either evolution did not happen or speciation can be observed around us on an ongoing basis like something out of the X-Men.

We are still being taught the 98% figure in upper-level university genetics courses.

In fairness, this information is relatively new, some of it from October 2018. Most professors probably don't even know about it yet.

Blogger Noah B. March 18, 2019 12:40 PM  

"Won't the TENS Supporters just expand the timeline again? Already its gone from a million years to a Billion Years. Now it will go to a Trillion years."

They can't. The current TENS timeline has been crafted to match findings from geology and astrophysics, so biologists don't have trillions of years to play with.

Blogger Jeff Burton March 18, 2019 12:43 PM  

It happened. It's a relevant number.

Oh come on. You know exactly what I meant. You thought that the mutation that increased mutations by 100X were good for you argument. But it isn't, because it just accelerates the destructive power of mutations.

Blogger peacefulposter March 18, 2019 12:48 PM  

Alright, the scientists might have been wrong about the whole evolution thing but they're correct about other important stuff, right? I mean, they're right about the sun causing skin cancer. Forget about the fact that very few pro golfers ever get skin cancer. Nothing to see there.

Wouldn't be surprised if those studies were funded by Coppertone.

Blogger Fargoth March 18, 2019 1:41 PM  

@40 And don't forget the most important thing scientists are really, really, right about! Vaccines!

Safe and effective! Safe and effective! Safe and effective!

Blogger maniacprovost March 18, 2019 2:57 PM  

attempting to conflate one of the underlying mechanisms by which genetic variation is generated with the differential survival/ reproduction of individuals possessing those mutations in a particular environment... into a single question.

The relevant questions are:
1. What is the spontaneous mutation rate per base pair per genome replication?

- see above Bionumbers summary.

2. What is the spontaneous mutation rate per base pair per generation (of the entire organism)?


No, guy. No. I kind of agree with you. I'm skeptical of Vox's argument. But no, he is not attempting to conflate two variables. He is attempting to use the variable rate of z=fixation/generation, which is itself a function of those two other variables. We don't need to know x=mutation/generation or y=fixation/mutation independently. We just need the end result.

My opinion is that we don't have enough data on population bottlenecking to get a good result. HOWEVER- I would like to see a mathematical work out of how population size affects fixation/generation.

Blogger steb March 18, 2019 4:01 PM  

Semi O/T: Hats off to Nature for publishing an article on the collapse of trust in science. It's section on climate change doesn't mention any of the past failed predictions and it doesn't acknowledge the reproducability crisis at all, but it does manage to discuss the holocaust and name-check Galileo an impressive 7 times.

'Science!' genuinely is a religion at this point.

https://www.nature.com/articles/d41586-019-00872-w

Blogger Kill Bill March 18, 2019 4:13 PM  

Unsurprisingly it was genetics and the Father of it, a Christian (Mendel) that deep 6'ed evilution.

Blogger Azimus March 18, 2019 4:17 PM  

38. Noah B The Savage Gardener March 18, 2019 12:40 PM
"Won't the TENS Supporters just expand the timeline again? Already its gone from a million years to a Billion Years. Now it will go to a Trillion years."

They can't. The current TENS timeline has been crafted to match findings from geology and astrophysics, so biologists don't have trillions of years to play with.


Have faith in science - they will find a way!

Blogger VD March 18, 2019 4:27 PM  

You're banned for lying and for refusing to answer a straightforward question, Francis.

I may have gotten a few things wrong about the e. coli studies, but you are an outright liar. Which, of course, I suspected on the basis of your first post which falsely claimed to address my argument.

It's going to be such a pleasure seeing you fetishists being forced to admit that skeptics like me were correct all along once the rest of the evolutionary biologists accept the inevitable.

Blogger SirHamster March 18, 2019 4:49 PM  

Gettimothy wrote:@14 I feeeeel like a banana.

Gregory the Great wrote:@16 And there we have the scientifically sound answer to why chimps are drawn to bananas.

You want banana? Be the banana.

Blogger tublecane March 18, 2019 5:47 PM  

The response to all this, mostly implied but sometimes spoken, will probably be "what else could it be?" And they have readymade insults for existing alternatives. Which is no way to do science, but old paradigms die hard.

I'm no longer certain ruling paradigms can shift anymore without violence.

Blogger Silent Draco March 18, 2019 6:21 PM  

It's the contamination and algorithmic construction issues that brought this to mind:

Back in the 1980s Tanker Wars in the Persian Gulf, the USS Vincennes accidentally shot down an Iranian airliner that made too close an axial approach. The Aegis air defense system interpreted it as an F-14 Tomcat on an attack profile, and recommended launch; airliner was shot down. Major diplomatic incident.

The why behind it was glossed over, with press reporting on stress for the captain and crew in a tense environment, robot warriors taking over, etc. Except - the Aegis radar system and algorithms were developed by RCA, in Cherry Hill, NJ. What did they initially train the algorithms on, as a test set? Conveniently, they were on the eastern approach axis to Philadelphia International Airport, with lots of Boeing and McDonnell-Douglas airliners in the traffic pattern. A large cross-section 'target' like an airliner should appear as a detection. So would an F-14 Tomcat, which, like an F-15, can be facetiously described as an exercise in corner reflectors (large radar return). Recognition and identification depend on what you're trained to detect. Sometimes, you get false recognition or ID.

"Dr. Tomkins: It has a huge effect because the chimpanzee genome is stitched together using the human genome as a scaffold. It’s like a puzzle—researchers used the human DNA “picture on the box” to assemble the chimp genome. The chimp DNA sequences used were all about 750 bases long. Not only was the chimp genome built using the human genome as a guide, it also has human DNA contamination in it, so it showed a lot of similarity from the contamination."

The test set was heavily human biased. Any tests on chimpanzee differentiation would identify a sample as highly similar to human. It was embedded in the test sets and algorithm structure.

The fundamental engineering and infrastructure get ignored, because media or science has another, shinier, answer. Every single time.

Blogger The Pitchfork Rebel March 18, 2019 6:25 PM  

"massive gap between the oft-reported 98 percent estimated similarity between the chimp and human genomes and what genetic scientists are actually seeing now as their ability to analyze the various genomes improves."

Let's assume for the sake of argument this is true. So what? What is common between humans and primates. Two eyes, one head, four limbs, live birth, are just basic architectural features. Where we differ-speech, cognition is what's important.

I'm not a geneticist, but an accountant and I know that if I produced two accounts, ledgers or anything else where the underlying detail was 98% in common-it would say nothing about the matter, because you can't draw any conclusions about the whole because 98% of the individual details are similar or even identical. You need to know the composition and effect of the differences.

I think Taleb's "Extremistan" is what's the important view.

Blogger Doktor Jeep March 18, 2019 6:31 PM  

We're not monkeys. News at 11. :-P

Blogger Lucas Evans March 18, 2019 6:35 PM  

Ok, why do we assume the chimp is our closest relative? If the similarity is only 80% then the chimp simply isn't very closely related, which makes sense.

Humans evolved from an arboreal ape based on our arm, hand and wrist structure. Chimpanzees are "knuckle walkers" and have a specialized wrists and hands. Humans are "palm walkers" when crawling on all fours.

So all this proves is that our original arboreal ape line died out. This happens all the time in nature. If they test enough extinct arboreal ape species, they might get a "hit" and find our original genetic line.

Blogger carry_bit March 18, 2019 6:51 PM  

Based on my work, I’m seeing not more than an 85% DNA similarity of chimpanzee to human, and that’s a maximum. It’s probably less than that.
If 1.5% is 9 million years, then 15% would be 90 million years. Maybe the Flintstones wasn't fiction after all!

Blogger tublecane March 18, 2019 7:25 PM  

@51- No joke, I think one main reason may be because newborn chimp faces remind people of human faces. And that we see ourselves in them more than other great apes.

Blogger MendoScot March 18, 2019 9:47 PM  

Small scale alignments are very informative.
But it doesn't scale up.
That men and male chimps, women and female chimps, were more related on % similarity than the intra-species genders should have been a heads up.

Blogger the SILLY explore March 18, 2019 11:55 PM  

This goes a little over my head as far as the details and their implications. Check out this summary.

The geneticists were probably sloppy with their assumptions, math, and lab procedure. The corrected data seem to mess up their preferred timeline so badly we can't even imagine how they'll make sense of it ("it" being the idea that humans descended from chimps or that they otherwise share a common ancestor). Of course this is not the only instance of geneticist making mistakes that all tend in one direction...

With these errors in mind we can be confident that the narrative in Genesis is more plausible than the idea that species evolved from a series of common ancestors. Even setting revelation aside, impartial consideration of the evidence points to the origin of mankind (and species in general, and even the entire universe) as having occurred roughly 10,000 years ago.

Am I on track? No one is mentioning Genesis, but that's why this is important right?

Blogger David F March 19, 2019 2:35 AM  

Richard Buggs' contention isn't exactly a fair comparison either. Of his figures, the SNPs are well aligned with the normally quoted 98-99%% figure. 4.06% comes from human DNA with no comparable sequence in the (less complete) chimp genome. 5.18% is in CNV (copy number variations).

He has already had to adjust his numbers upward as sequencing improves, so his assumption that the 'unknown' 4% will all end up with no comparable sequence isn't a good one. It would be much fairer to assume that it will follow the pre-existing trend of the other 96%. So broadly, we can discard it.

The CNVs are interesting - repeated sections of DNA that show up more or less often in one genome or the other, but are still fundamentally similar. They are obviously a different class of transcription error from the SNP (single nucleotide polymorphism - basically a point error that one can view as an isolated "bit flip" error), where large blocks of DNA get inserted or deleted.

To claim that the Chimp genome is 5% different from the human genome due to CNVs is to some extent comparing apples to oranges. If chimps have 6 copies of sequence A and humans only have 4 copies, is it really fair to add 2*(length of A) to the list of differences? This is where we're running into a difference of definitions, not a difference of fact.

The genetic drift clock only looks at SNPs, not at the big insertion/deletion/transcription events. Buggs' figures show 1.18% of the genome different due to SNPs. 3x10^9 base pairs in the human genome * 0.0118 = 35 million SNPs.

So Buggs isn't helping your "time to chimp human last ancestor" argument at all.

Nor is the fact that you did not account for 5 orders of magnitude difference in per-generation mutation rate between the bacteria of your Nature 2009 study and humans. Fixation rate is proportional to mutation rate, so those five orders of magnitude bumps your 281 'maximum mutations' to 28 million.

(To reiterate: 1000x from differences in genome length, 100x from average number of cell replications from fertilized egg to next generation's egg and sperm.)

Blogger Tom Bridgeland March 19, 2019 4:36 AM  

Anyone have access to a chimp? Be very intersting to get a swab and submit it to 23andMe.

Blogger SirHamster March 19, 2019 3:20 PM  

David Fenger wrote:It would be much fairer to assume that it will follow the pre-existing trend of the other 96%. So broadly, we can discard it.

"fairer to assume"

"Let's assume we know something about something we don't know."

Garbage goes in, garbage comes back out.


David Fenger wrote:So Buggs isn't helping your "time to chimp human last ancestor" argument at all.

Vox's argument is that a number for "time to chimp human last ancestor" exists and can be found if evolution is true. Furthermore, that this number needs to be within a certain range for TENS to hold true.

Your arguing against the number Vox calculated doesn't address or hurt Vox's argument because you missed the point.

That I need to spell this out for you is a hint that you are not tall enough for the ride.

Blogger David F March 19, 2019 3:56 PM  

@58 The TENS calculation for Vox's CHLA problem is very different from his calculation. The TENS approach includes what I stated: fixation rate is proportional to mutation rate.

The standard TENS calculation for Vox's CHLA problem comes up with 28 million mutations in the human branch, not 281.

Blogger SirHamster March 19, 2019 4:27 PM  

David Fenger wrote:The TENS calculation for Vox's CHLA problem is very different from his calculation.

That doesn't undermine his argument. Now you're bargaining over the numbers that his argument argues exists.

David Fenger wrote:The TENS approach includes what I stated: fixation rate is proportional to mutation rate.

That is assumed, not observed. The lack of observation and testing makes that not science.

David Fenger wrote:The standard TENS calculation for Vox's CHLA problem comes up with 28 million mutations in the human branch, not 281.

28 million mutations over 12 million years averages 2 mutations a year, 40 per 20 year generation.

What are the 100 mutations fixed into the human genome over the past 50 years?

Blogger Ryu238 March 19, 2019 8:30 PM  

"The first exhibit is an interview with a creationist geneticist, which will no doubt be improperly dismissed by scientistry fetishists with an appeal to the genetic fallacy." No we show how often this 'scientist'lies. https://pandasthumb.org/archives/2013/06/do-the-creation.html

Blogger David F March 20, 2019 2:49 PM  

@60 The correlation between mutation rate and fixation rate is fairly easy to show from first principles. I even wrote a tiny little simulation that makes the point, linked in the previous discussion on this.

Agreed, the fixation rate is somewhere in the 30-70 rate per generation, same as the mutation rate.

I have no idea what the 100 mutations fixed in the last 50 years are. Nobody does, and it's likely we never will. You'd have to measure the genome of everyone on the planet, every generation.

I can tell you a little bit about them, though. They're SNPs (individual base-pair errors) somewhere that has no impact on viability. The mutation would have happened many generations ago, as it takes a long time to random-walk to fixation. The last person with the 'unmutated' site is not obviously different from their neighbours. They could have had children, but due to the luck of the draw, the chromosome with the site in question lost the coin flip for each child, and their cells are the last that have the 'unmutated' site.

And when this one person in 70 million dies, they take that linage with them, with no fanfare.

Blogger SirHamster March 20, 2019 5:57 PM  

David Fenger wrote:The correlation between mutation rate and fixation rate is fairly easy to show from first principles.

I heard you the first time, retard. No need to repeat yourself like a broken record.


David Fenger wrote:I have no idea what the 100 mutations fixed in the last 50 years are. Nobody does, and it's likely we never will.

You're a fraud and you know it.

"Mutations are getting fixed in the population over time, but it's impossible for us to detect and measure those mutations."

If the mutation is fixed, it's in the gene pool where it was not before. A random sampling of individuals each generation would provide snapshots that can be compared.

Maybe it's too small a change to easily and reliably detect over 1 generation, but you should be able to detect the fixation of 1000 mutations, or maybe even 100,000 mutations.

"never"? You know your model is junk, and discourage the use of observation to validate the model. Your model can't be validated by science.

You just declared your model, the "standard TENS calculation," unscientific. I already know that, but hearing it from you was worth the price of admission.

Blogger David F March 20, 2019 7:46 PM  

@63 And back to ad hominem. Fun.

If it's so simple, show me where the math fails.

We can detect the individual mutations from parent to child with effort - but the effort is significant, so the experiment hasn't been done many times.

To detect the fixation of thousands of mutations over time we'd need to have samples from thousands of years ago to compare to present DNA. We don't have those.

What we do have is divergent lineages. Us vs chimps, chimps vs bonobos, all three vs orangutans, and so on. In each case, the number of SNPs between the genomes correlates passably well to the time since their divergence in the fossil record, by the calculations I've given above.

If there was a fundamental problem with the gene clock, why does that correlation exist?

SirHamster wrote:Your model can't be validated by science.

All I've claimed is that the particular observation you've asked for is infeasible. I can't show you fixation in the human population, it's too vast, and too difficult to measure. That one particular measurement can't be made doesn't invalidate others.

We can measure the mutation rate in humans.

We can measure the effects of varying mutation rate on fixation rate, and see that it confirms the theories about rate of genetic drift.

We can measure genetic drift between vast numbers of pairs of species, and find again and again and again that the gene clock correlates reasonably well with the fossil record.

What theory do you advance to explain the above?

Blogger SirHamster March 20, 2019 8:13 PM  

David Fenger wrote:And back to ad hominem. Fun.

Quit acting like a retard and you won't get called one.


David Fenger wrote:

We can detect the individual mutations from parent to child with effort - but the effort is significant, so the experiment hasn't been done many times.


David Fenger wrote:All I've claimed is that the particular observation you've asked for is infeasible. I can't show you fixation in the human population, it's too vast, and too difficult to measure. That one particular measurement can't be made doesn't invalidate others.

Two claims here:

- You can detect individual mutations in one generation.

- You cannot detect the fixation of mutations in humans over 50 years.


One of these claims is false. You're a fraud and you know it.

All you had to do was tell the truth: "It's impractical to detect the 100 mutations my model predicts." But that brings uncomfortable followup questions.

So instead, you lie that it can't be done and never will. That's why I call you a retard - you're making stupid lies to protect your position.

Blogger David F March 20, 2019 9:03 PM  

@65 Detecting individual mutations in one generation requires sequencing 3 people - mother, father, child.

Detecting fixation of mutations in the population requires sequencing *everyone*. And then doing it again next generation.

The cost of sequencing a human's genome is currently around $1000. Finding the mutation rate costs about $3000. Not hard to find the budget for that.

The cost of sequencing *all of humanity* just once is $7 trillion. It could theoretically be done, but the effort is such that I cannot imagine it happening until gene-sequencing becomes a staple of medical care. Even then, privacy laws would prevent getting the full sample set.

How is "It's impractical to detect the 100 mutations you're asking for" different from what I said? "Nobody does, and it's likely we never will" is somewhat more florid, I grant. But I agree with your statement above.

To state it more clearly:
* The measurement of the current fixation rate over all humanity over the last 50 years has not been done.
* The cost of making the measurement at the current technological level makes it unlikely anyone will try.
* There may come a time when sequencing is cheap and common enough that we can get a large enough data set to make a good estimate from it. Given the way the costs have levelled out, and the increasing concern with privacy, that may never happen.

What are your follow up questions?

Blogger SirHamster March 20, 2019 10:07 PM  

David Fenger wrote:What are your follow up questions?

Why did you lie that "impractical" is "can't"?

Why wouldn't you describe the method for verifying that your model is correct, even if it hasn't been done, and is impractical to implement in the near future?

Why are you discouraging investigation and confirmation that your model is true?

Blogger David F March 20, 2019 11:19 PM  

SirHamster wrote:Why did you lie that "impractical" is "can't"?

Echoing your use of it, I think. I did not say it can't be measured, I said I can't show it to you, and used again as a general statement: "That one particular measurement can't be made doesn't invalidate others." I'll grant it's imprecise. A more accurate statement would be to substitute "can't" for "hasn't", same general point.

SirHamster wrote:Why wouldn't you describe the method for verifying that your model is correct, even if it hasn't been done, and is impractical to implement in the near future?

Because I don't need to measure fixation across the entire human species to verify the model. You're the one asking for that. We can measure fixation in model species that can be studied in more reasonable amounts of time and effort, then make a reasonable effort at projecting those results to the target species.

This is what Vox did with the CHLA example. The model species is bacteria, the target is humans and chimps. The 40,000 generations of bacteria took about 20 years to fix 25 mutations, so the net fixation rate was about 1.25 per year.

Does that mean 9 million years of human evolution is expected to produce 11 million mutations? There are a few differences between humans and bacteria after all, so we need to apply some translation factors.

Vox got one of them: generation time. The bacteria got 2,000 generations per year, humans get one per 20, a scaling factor of 40,000. So instead of 11 million mutations, we're down to 275. I've done some rounding here, so I'm off by a bit, but that's close enough to the 281 Vox should have got if he hadn't swapped between 4 million and 9 million years halfway through.

He missed other significant factors. Generation time is the wrong scaling term, because human cells replicate on average 100 times per generation. So that gets us from 275 to 27,500 mutations.

The other one is genome size. The human genome is around 3x10^9 base pairs, the bacteria 5x10^6. That's a scaling factor of 600. 27,500 * 600 = 16.5 million fixations.

The genome-size/mutation rate scaling factor is controversial here, but it is a broadly accepted and well documented aspect of TENS.

SirHamster wrote:Why are you discouraging investigation and confirmation that your model is true?

I have no wish to discourage investigation into the model. In previous threads I've provided a variety of links to the sources for my data. A good starting point:
http://book.bionumbers.org/what-is-the-mutation-rate-during-genome-replication/

The above article gives the genome sizes for bacteria and humans, links the Nature 2009 article Vox cited and another on direct measurement of human mutation rates. I got the 20 year figure for the 40,000 generations from https://www.nature.com/articles/4611219a


Blogger SirHamster March 21, 2019 1:02 AM  

David Fenger wrote:SirHamster wrote:Why did you lie that "impractical" is "can't"?

Echoing your use of it, I think. I did not say it can't be measured, I said I can't show it to you, and used again as a general statement


You are a liar. I did not prompt your "can't" claim.

David Fenger wrote:I have no idea what the 100 mutations fixed in the last 50 years are. Nobody does, and it's likely we never will.

You said it will never be known.


David Fenger wrote:Because I don't need to measure fixation across the entire human species to verify the model.

Of course the fraud doesn't believe in verifying the model. We're supposed to believe that humans have high mutation fixation rates without ever observing it.

David Fenger wrote:I have no wish to discourage investigation into the model.

Saying it will never be known is saying investigation is pointless. It's also placing your claims outside the realm of science.

Which would be fine if you fraudsters would also stop pretending to science.

Blogger David F March 21, 2019 1:58 AM  

@69 I said "it's likely we never will". I personally do not think it likely that that measurement will ever be done.

SirHamster wrote:Of course the fraud doesn't believe in verifying the model. We're supposed to believe that humans have high mutation fixation rates without ever observing it.

I believe in verifying the model. I also happen to believe the model has been robustly verified with the measurements done to date. We have radically different opinions on what it would take to verify the model.

What I considered when trying to validate the model:
A) Fixation of mutation at a rate proportional to the observed neutral mutation rate has been observed. Granted, most of that work has been done on bacteria, but their DNA works like ours does.
B) There is a simple and elegant mathematical explanation of the relationship. N*u * (1/N) = u, see the wikipedia page on fixation for details.

However, I had a hard time wrapping my head around (B), so I took one extra step:
C) I created and examined a very simple computer model of mutation and fixation, and it came up with the same result as an emergent behaviour.

The calculation of the expected number of mutations over 450,000 generations of humans relies on the relationship between mutation rate (which has already been measured), and the fixation rate (which as discussed above, requires immense effort).

SirHamster wrote:Saying it will never be known is saying investigation is pointless. It's also placing your claims outside the realm of science.

That one measurement is too difficult to attempt does not stop the process of science. It just means one has to find another way to approach the problem. If one can't measure it directly, find a way to do it indirectly.

SirHamster wrote:Which would be fine if you fraudsters would also stop pretending to science.

I am in no way attempting to pull the wool over your eyes, rather the opposite. Science observes what it can, builds hypotheses, turns those into models, and attempts to find new observations that confirm or deny those hypotheses. When the models work well enough, accurately predicting measurements before they are made, the models are treated as accurate and useful... until someone finds an edge case or observation that breaks the models. Then the process starts over. Most of the time, the model gets refined with more terms. Sometimes the model grows overly complex, and a new, simpler one takes its place.

An example of the test and confirmation process is here: https://www.nature.com/articles/nature11128

It is a study making a new measurement (at that point in time, 2012), sequencing a bonobo's genes to see how they compare with chimp and human genomes. The number of SNP differences between bonobos, chimps, and humans all correspond well to the time since divergence in the fossil record.

Thus the model is confirmed, data point by data point. There are a *lot* of data points like that.

Blogger Wild Man March 21, 2019 2:20 PM  

@56 - David Fenger - you said - "The CNVs are interesting".

Here is an interesting article about types of viruses for which the genes are distributed among distinct genetic-packet varieties (each with distinct protein coding domains), and perhaps (as the article speculates) this type of viral condition might offer adaptation benefits, by way of a CNV (copy number variation) strategy that would otherwise have not become an adaptive strategy, if CNV occurrence was limited to only operate non-distributively in these types of weird viruses (because of the physical/energetic limitations with respect to genome size for viruses):

https://www.theatlantic.com/science/archive/2019/03/the-revolutionary-discovery-of-a-distributed-virus/584884/

What is not clear to me is - do we know the physical operation by which CNV type coding changes occurs in higher life forms? If not - perhaps there is a hint available, by way of these strange types of viruses.

David - I can see that you are in the best position here, expertise-wise, to weigh in on these issues in a credible way - so ...... that is why I am asking you. I hope you may see my comment before VD removes it. Do you have a website of your own where I could engage with you, without VD's annoying message-policing?

Blogger SirHamster March 21, 2019 6:00 PM  

David Fenger wrote:I also happen to believe the model has been robustly verified with the measurements done to date.

It isn't, and no one here is dumb enough to believe that your mathematical model is a proper substitute for experimentation.

We're not asking you to rationalize your beliefs to yourself. We're asking for a level of evidence that is persuasive to skeptics, as skeptics.

Meeting that standard is optional, but your use of lies and fraud demonstrate your opinion on the subject is worthless.


David Fenger wrote:That one measurement is too difficult to attempt does not stop the process of science.

Again, you're a fraud. The measurement you pointed to is 3 samples: father/mother/child. That covers 2 generations.

You could do an additional 4 samples of the grandparents to cover 3 generations, which is 60+ years of mutations within a single family branch.

7 samples to measure a single branch's fixation. Then another 7 samples to measure a second branch. A fixed mutation will show up in both branches. It's not a binary option between measuring 0 samples and measuring the entire world.

That's an additional 11 samples to do a quick verification of your claim of fixation rates. Let's call that 3 extra studies worth of work.

1 study is enough for you to claim you know what the human mutation rate is.

But doing the equivalent of 3 extra studies to measure and verify fixation rate is according to you: "too difficult", "infeasible", and "never will" be done. BS.

Pretty much anyone responding to you has at least a college degree. We know how to do science. We can also tell what isn't science.

What you're doing is not science. It is fraud.

And we all know it's fraud, because even you know if we measured the mutation fixation in 2 randomly selected branches of grandparent/parent/child, those mutations won't match, and they are not fixed by definition. There is no path to global gene transfer over 3 generations.

That's why I'm describing the experiment, while you're cowardly avoiding it - even as you fraudulently claim your model is "robustly verified".


David Fenger wrote:I am in no way attempting to pull the wool over your eyes, rather the opposite.

You have repeatedly lied about the possibility of verifying your model, and this is another lie.

You're a fraud, you've been called out for fraud, you've doubled down rather than retract, and no one is fooled.

Blogger David F March 22, 2019 2:50 PM  

@71 - Wild Man - That is a fascinating find. It's like the virus lives with continually varying CNVs, in some sense. I'd be a little reluctant to accept that it behaves the way they describe without some corroboration. It sounds more like co-infection than a single virus split amount 8 packets, each capable of reproduction.

I'm afraid I'm no geneticist, though. Just an interested amateur. I'm more of a mathematician and programmer.

My (limited) understanding of CNVs is that the DNA replication process is somewhat fragile - the strand breaks every now and then, and there are mechanisms in place to splice it back together when that happens. The thing is, if the break is in a section of DNA where the same pattern repeats over and over, it's easy to end up splicing at the wrong spot in the pattern.
So instead of 20 copies of of the sequence, now you have 18, or 22.

Don't take my word on it, though. For all it's maligned here, wikipedia is a decent starting point, and following the references in the articles will lead you to much deeper articles on the subject. (Start at Mutation, subheader Errors introduced during DNA repair.)

I don't have my own website as such, but would be amenable to chat via gmail.

Blogger David F March 22, 2019 3:20 PM  

@72 There has been plenty of experimentation done, and I've linked to a variety of papers doing just that.

Tell me what would meet your standard.

SirHamster wrote:You could do an additional 4 samples of the grandparents to cover 3 generations, which is 60+ years of mutations within a single family branch.

This is a useful measurement, and one commenter suggested that it had been done. I haven't managed to find a paper describing it, so I'm relying only on what I've found to date, which was based on a pair of parent-parent-child studies (two different families), and produced a mutation estimate in the 10-100 range. Huge error bars, due to paucity of data points. It was an older paper, so I expect there are better estimates now. I haven't looked that hard, and haven't seen them yet.

However, this does not measure *fixation* rates. You wanted to proof of fixation in humans. Fixation is a phenomoenon of populations - a mutation is not fixed in the population until its competitor is extinct.

SirHamster wrote:Pretty much anyone responding to you has at least a college degree. We know how to do science. We can also tell what isn't science.

I'm not doing science. I have no budget to do so with, nor a background in genetics. I'm reporting on the results of scientists, and trying to get the math right. Mathematics and programming is my background. And I have more than one degree to my name.

SirHamster wrote:you know if we measured the mutation fixation in 2 randomly selected branches of grandparent/parent/child, those mutations won't match, and they are not fixed by definition. There is no path to global gene transfer over 3 generations.

That's one heck of a straw man there. I have made no such claim. In fact, you're arguing against your points above.

Today's mutations aren't going to get fixed any time soon. The ones found via any given grandparent-child sequencing experiment aren't even likely to fix at all. The odds are vastly against it for any given mutation - 1/N.

It's not this generation's mutations that fix. Or last generation's. Or even 100 generations ago. It's mutations from thousands or more generations ago. Mutations that are already present in 99.9999% of the population, and are just waiting for the right people to die before they stand alone.

SirHamster wrote:That's why I'm describing the experiment, while you're cowardly avoiding it - even as you fraudulently claim your model is "robustly verified".

The experiment you propose above does not show the fixation rate, only the mutation rate. Which is great, we want to know that... but that experiment doesn't answer your original question. To quote:
SirHamster wrote:What are the 100 mutations fixed into the human genome over the past 50 years?

What experiment do you propose that would give you your answer?

Blogger SirHamster March 22, 2019 5:55 PM  

David Fenger wrote:SirHamster wrote:You could do an additional 4 samples of the grandparents to cover 3 generations, which is 60+ years of mutations within a single family branch.
...
However, this does not measure *fixation* rates. You wanted to proof of fixation in humans. Fixation is a phenomoenon of populations - a mutation is not fixed in the population until its competitor is extinct.


The experiment measures 2 branches, not 1. It measures a population of 2 grandchildren and compares it to a population of 8 grandparents.

A better experiment could be designed. This was done quick, it didn't take 7 trillion dollars or 100% sampling, and it has the potential to find 100 fixed mutations that the grandchildren have and the grandparents don't.

Not looking for proof of fixation, just an actual measurement of what you claim is taking place.


Anyone who treats an un-validated computer model as a substitute for experiment is engaging in fraud.

Blogger David F March 23, 2019 2:41 AM  

@75 You do not appear to understand what a fixed mutation is.

"In population genetics, fixation is the change in a gene pool from a situation where there exists at least two variants of a particular gene (allele) in a given population to a situation where only one of the alleles remains."

Measuring a handful of people can tell us the mutation rate. But to tell that a given allele has vanished from the gene pool, we have to look for it everywhere.

SirHamster wrote:Not looking for proof of fixation, just an actual measurement of what you claim is taking place.

I can't show you fixation in the whole of the human race over short time scales. The evidence over long time scales is exactly what we're arguing over - interpretation of the 30 million SNPs between human and chimp genomes.

I've linked a few examples of studies showing fixation in bacteria. A good one:
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4503671/

What would it take to convince you that the relationship between mutation rate and fixation rate is fairly well understood?


SirHamster wrote:Anyone who treats an un-validated computer model as a substitute for experiment is engaging in fraud.

The computer model hasn't been scrutinized, but I did offer up the source code for critique. Here it is again: https://pastebin.com/cgBRwyMc

Grab tinycc or any other C compiler of your choice, and have at it.

It is just about as simple as I could possibly make it. Digital organisms with large genomes (within an order of magnitude of the bacteria from the original study) replicate, and each generation lives or dies with a 50% probability, modified by any negative mutations they've acquired. There's no prior assumptions about fixation baked into it, all it does is try to measure the fixation rate over generations of constant population. It's not a "hide the decline" climate model or anything like that, and not even trying to represent any specific organism.

All I wanted to do was use it to understand the correlation between mutation rate and fixation rate, because u*N * 1/N = u was more than a little counter-intuitive. That, and writing little programs like that is fun for me.

Blogger SirHamster March 24, 2019 12:44 AM  

David Fenger wrote:Measuring a handful of people can tell us the mutation rate. But to tell that a given allele has vanished from the gene pool, we have to look for it everywhere.

No one asked you to find vanished alleles.

When 2 samples share the same allele, is it fixed?

How about 3? 5? 10? 100?

Did your multiple degrees of education not cover measurement error?

David Fenger wrote:What would it take to convince you that the relationship between mutation rate and fixation rate is fairly well understood?

You would have described an experiment that measures fixation rates and demonstrated the relationship, instead of lying that "we never will".

Note the past tense. Your repeated appeals to computer models demonstrates that the relationship is not well understood.

Blogger David F March 24, 2019 10:41 PM  

SirHamster wrote:When 2 samples share the same allele, is it fixed?

How about 3? 5? 10? 100?



No, no, no and no.

If 100 well-distributed measurements show it as fixed, I'd be willing to concede that the allele is 'likely fixed', but with a margin of error that I'd not be entirely happy with. It wouldn't help us find which allele fixed in the last generation. That needle requires a much bigger haystack.

SirHamster wrote:You would have described an experiment that measures fixation rates and demonstrated the relationship

Which I have. See: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4503671/

SirHamster wrote:instead of lying that "we never will".

My statement was "I have no idea what the 100 mutations fixed in the last 50 years are. Nobody does, and it's likely we never will." I stand by it - we'd have to have genetic material archived from everyone who died in the past 50 years to make the complete measurement.

That impossible goal does not prevent us from making estimates of the rate at which fixation can reasonably be expected to occur over hundreds of thousands of generations.

The computer model is of limited use, I grant. It does put paid to Vox's complaints about parallel versus serial fixation, by making it possible to see how parallel fixation works step by step if one is so inclined.

The stronger argument is simple math: u*N * 1/N = u.

Post a Comment

Rules of the blog
Please do not comment as "Anonymous". Comments by "Anonymous" will be spammed.

<< Home

Newer Posts Older Posts