From the title through the posthumous shoutouts, a great piece of writing. Thanks for illuminating these percentage differences. I've hankered to understand that for a while; your explanation cleared it right up for me.
Love your substack - your writing is excellent and you have fueled a real curiosity in me to learn more. Wondering if you might provide a recommendation of a good genetics primer for a someone who is willing to put in the time to understand genetics more fully. I can follow along with your writing but a next level understanding would do well for me. Any ideas would be welcome - and if other readers have found good references for genetics understanding I would be please to those as well!!
Thank You for this well understandable summary ! Some 20 years back I saw an author write of Homo sapiens as "the human animal". I felt this embarassing, yet I wanted to understand why. For many years I was surprised to learn how much great Apes behave like 'us'. Now, when considering our human dealing with our natural environment and amongst ourselves, my view is changing: I more and more see how much "we" still behave like "them". Obviously it's a very looong way for us to outgrow our heritage... or to accumulate enough 'fixed mutations' to no longer kill ourselves so easily. I suppose the next version of hominin will do so.... at least some of the survivors... I'm an optimist, am I not ?
Rousing piece and I strongly endorse the sentiment of striving to find more simply because we can. I have a question regarding the 23andMe shared segments point you made.
In your haplogroup piece you mentioned how the chromosome I receive from each parent is actually a mish-mash of the two chromosomes that parent has at that location due to genetic recombination during meiosis.
My question is, how then can 23andMe claim that I share the entire chromosome at precisely the same locations with my parent when those locations have been messed up by the genetic recombination ?
good point! i didn't want to dig too deeply, but basically what they do is just assign BOTH the contributed copies to a given parent. so, chromosome 1 that i contribute to my son is a mix of my father and mother along recombination points. so it doesn't 'match' a single chromosome 1 of mine...but, it matches a synthetic one.
I would be interested to hear from you on as to what types of genes most commonly differ, and by how much (i.e , between us and chimps)? Genes related to:
Immune function,
Muscle fiber strength,
BRCA and other DNA repair genes,
Mylen sheath composition genes,
O2 uptake by hemoglobin as a function of pH,
Etc.
I.e. is it necessarily things we might expect to be most different? and What are the real surprises?
How comparable are the differences between us and chimps, us and Bonobos, and chimps and Bonobos... Etc.
So... you’re saying there is a chance a part of Homo naledi lives inside us!
Other people have already said it, but this is beautiful. Sincerely, this is spiritual. We do owe it to those who came before us to tell their stories to the best of our abilities. I don’t understand how everyone is not obsessed this stuff. I’m glad I reread this tonight. Another thing that’s spiritual, and pretty trippy, is being able to compare your DNA with your sons DNA. That must have been a really cool experience for you to have.
Your explanation of Identity by Descent--defined as a matter of “tracking genetic ancestry within shallow genealogies or pedigrees”--seems unsatisfactory. The shallowest genealogy is one generation, which would have one sharing roughly half one’s genes with a parent but none with a grandparent. But the actual assumption is that one shares roughly ¼ with a grandparent, 1/8 with a great-grandparent, etc., indefinitely far back (though petering out eventually because the number of genes is finite). No one relies on a concept based on “shallow genealogies.”
There’s a great mystery about chromosome number that I have posed in blog comments to several geneticists and never gotten a serious answer. How does the number of chromosomes change in a way that gets fixed within a group without automatically causing both speciation and a very tight genetic bottleneck?
Are there any species which are recognized as a single interbreeding species with a common gene pool, where the haploid number varies as it does between chimps and humans?
If so, is that the general case for situations like the chimp-human divergence, where maybe differential cross-fertility leads to a natural sorting into groups that quickly lose the ability to interbreed, perhaps because chromosome fusion-fission sites diverge too much?
If not, then how is an extreme bottleneck avoided?
"Are there any species which are recognized as a single interbreeding species with a common gene pool, where the haploid number varies as it does between chimps and humans?"
horse.
geneticist who work in this area tell me that the chromosome number variation as preventing interbreeding is overdone. i don't know myself much about this
That doesn’t explain why when the common ancestor of humans and chimps forked, one group settled on 23 pairs and the other on 24 pairs. Do we have any idea how to tell whether the ancestral ape had 23 pairs, 24 pairs, or a mixture? Is it possible to tell the difference between whether the two chimp pairs and the one human pair came from a fusion or a splitting? Do gorillas have an analogue of the one human chromosome, two analogues of the two chimp chromosomes, or something more complicated at those alleles?
gorilla are 24. so the assumption is the ancestral is 24.
there is a literature in chromosome evolution. i'm not super versed in it but you can dig into it (most of it is in drosophila and has to do with sex chromosomes).
That’s not too relevant. I care about mammalian autosomes and want to know the data on that. If the fork between chimps and humans was before, coincident with, or after the 48->46 fusion seems like a really important trichotomy!
Before: how did 48 become 46 without a tremendous bottleneck?
After: How did chimps go from mixed 48-46 to only 48?
During: maybe this is the main way speciations actually happen in mammals?
"Before: how did 48 become 46 without a tremendous bottleneck?"
ok, i'll connect the dots here since i assumed you would see what i was saying: mongolian wild horses and domestic horses breed and produce fertile offspring despite different chromosome numbers. your assumption that this requires a massive bottleneck is just false.
" If the fork between chimps and humans was before, coincident with, or after the 48->46 fusion seems like a really important trichotomy!"
why? we're more close to chimps than gorillas. you place bit too much emphasize on karyotype
The reason I think it’s important is that either a particular difference in chromosome structures is no significant bar to interbreeding, a major handicap to it, or rules it out entirely. Each of the three hypotheses has problems!
If no significant bar, why did both humans and chimps fixate on different numbers, after the fork?
If it is a major handicap, how did the common ancestor of humans and chimps manage to have such a mix? If it is ruled out entirely, how does a divergence arise without an extreme bottleneck?
If there were just ONE pair of now-mutually-infertile mammalian species differing in this way where what actually happened had been traced and figured out, I would be a lot less puzzled.
I used to think that the 48 vs 46 human chromosome count was a shocking example of scientific groupthink, but I recently looked more carefully at the history, and maybe not quite. My reading is that mainly it was that until the 1950s basically nobody really cared about the number, though indeed one guy in the 1920s reported the human number as 48. It appeared in textbooks, so outsiders assumed it was True, but the (known to insiders) somewhat crappy original estimate was left alone mainly because it didn't seem an important question.
right. basically ibs is simple: does the base or variation match? the provenance is irrelevant. 99.9% of a given human's genome matches another human. why? not because they are closely related in any colloquial sense, but because we're a young species that share common ancestors within the last 300,000 years.
ibd is more recent shared ancestry. go back far enough and it's all the same ancestors, but in the intervening time there has been lots of geographic separation, endogamy, drift and divergence. so traversing the recent pedigrees and genealogy is what looking at 'ibd segments' does. it exposes connections along the last 50 generations or so. really the last 10 in most cases.
You mention that “an example would be the gene that he possesses in common with Europeans: not because of common descent, but because of selection for that gene resulting in separate populations having that gene.” But most likely this gene *has* come down from a common ancestor of him and the Europeans (maybe as long as 4,500 years ago). It *might* have arisen by mutation independently in separate populations, but single origin and common descent therefrom is usually more likely. (Of course, the gene can be selected for in different populations even if it has a single origin in an ancestral population.)
" It *might* have arisen by mutation independently in separate populations, but single origin and common descent therefrom is usually more likely."
you are right about the haplotype, but not necessarily about the SNP. in most cases these concepts arose in the context of a specific marker and not a genomic segment (genomics did not exist)
From the title through the posthumous shoutouts, a great piece of writing. Thanks for illuminating these percentage differences. I've hankered to understand that for a while; your explanation cleared it right up for me.
Inspiring
Hi Razib,
Love your substack - your writing is excellent and you have fueled a real curiosity in me to learn more. Wondering if you might provide a recommendation of a good genetics primer for a someone who is willing to put in the time to understand genetics more fully. I can follow along with your writing but a next level understanding would do well for me. Any ideas would be welcome - and if other readers have found good references for genetics understanding I would be please to those as well!!
perhaps the next 'time well spent'
Thank You for this well understandable summary ! Some 20 years back I saw an author write of Homo sapiens as "the human animal". I felt this embarassing, yet I wanted to understand why. For many years I was surprised to learn how much great Apes behave like 'us'. Now, when considering our human dealing with our natural environment and amongst ourselves, my view is changing: I more and more see how much "we" still behave like "them". Obviously it's a very looong way for us to outgrow our heritage... or to accumulate enough 'fixed mutations' to no longer kill ourselves so easily. I suppose the next version of hominin will do so.... at least some of the survivors... I'm an optimist, am I not ?
well steven pinker thinks we're getting less violent. so there's that :)
Rousing piece and I strongly endorse the sentiment of striving to find more simply because we can. I have a question regarding the 23andMe shared segments point you made.
In your haplogroup piece you mentioned how the chromosome I receive from each parent is actually a mish-mash of the two chromosomes that parent has at that location due to genetic recombination during meiosis.
My question is, how then can 23andMe claim that I share the entire chromosome at precisely the same locations with my parent when those locations have been messed up by the genetic recombination ?
Thanks
good point! i didn't want to dig too deeply, but basically what they do is just assign BOTH the contributed copies to a given parent. so, chromosome 1 that i contribute to my son is a mix of my father and mother along recombination points. so it doesn't 'match' a single chromosome 1 of mine...but, it matches a synthetic one.
Thanks Razib that clarifies things. And do they do the same for more distant relatives say, 3rd cousins too ?
so more than 3rd cousin last i checked was mostly ppl you are not 'really' related to...you just share ibd segments floating in the population
I would be interested to hear from you on as to what types of genes most commonly differ, and by how much (i.e , between us and chimps)? Genes related to:
Immune function,
Muscle fiber strength,
BRCA and other DNA repair genes,
Mylen sheath composition genes,
O2 uptake by hemoglobin as a function of pH,
Etc.
I.e. is it necessarily things we might expect to be most different? and What are the real surprises?
How comparable are the differences between us and chimps, us and Bonobos, and chimps and Bonobos... Etc.
So... you’re saying there is a chance a part of Homo naledi lives inside us!
Other people have already said it, but this is beautiful. Sincerely, this is spiritual. We do owe it to those who came before us to tell their stories to the best of our abilities. I don’t understand how everyone is not obsessed this stuff. I’m glad I reread this tonight. Another thing that’s spiritual, and pretty trippy, is being able to compare your DNA with your sons DNA. That must have been a really cool experience for you to have.
Your explanation of Identity by Descent--defined as a matter of “tracking genetic ancestry within shallow genealogies or pedigrees”--seems unsatisfactory. The shallowest genealogy is one generation, which would have one sharing roughly half one’s genes with a parent but none with a grandparent. But the actual assumption is that one shares roughly ¼ with a grandparent, 1/8 with a great-grandparent, etc., indefinitely far back (though petering out eventually because the number of genes is finite). No one relies on a concept based on “shallow genealogies.”
wrong. ask any forensic scientist
This was a fun read Razib
There’s a great mystery about chromosome number that I have posed in blog comments to several geneticists and never gotten a serious answer. How does the number of chromosomes change in a way that gets fixed within a group without automatically causing both speciation and a very tight genetic bottleneck?
Are there any species which are recognized as a single interbreeding species with a common gene pool, where the haploid number varies as it does between chimps and humans?
If so, is that the general case for situations like the chimp-human divergence, where maybe differential cross-fertility leads to a natural sorting into groups that quickly lose the ability to interbreed, perhaps because chromosome fusion-fission sites diverge too much?
If not, then how is an extreme bottleneck avoided?
"Are there any species which are recognized as a single interbreeding species with a common gene pool, where the haploid number varies as it does between chimps and humans?"
horse.
geneticist who work in this area tell me that the chromosome number variation as preventing interbreeding is overdone. i don't know myself much about this
That doesn’t explain why when the common ancestor of humans and chimps forked, one group settled on 23 pairs and the other on 24 pairs. Do we have any idea how to tell whether the ancestral ape had 23 pairs, 24 pairs, or a mixture? Is it possible to tell the difference between whether the two chimp pairs and the one human pair came from a fusion or a splitting? Do gorillas have an analogue of the one human chromosome, two analogues of the two chimp chromosomes, or something more complicated at those alleles?
gorilla are 24. so the assumption is the ancestral is 24.
there is a literature in chromosome evolution. i'm not super versed in it but you can dig into it (most of it is in drosophila and has to do with sex chromosomes).
That’s not too relevant. I care about mammalian autosomes and want to know the data on that. If the fork between chimps and humans was before, coincident with, or after the 48->46 fusion seems like a really important trichotomy!
Before: how did 48 become 46 without a tremendous bottleneck?
After: How did chimps go from mixed 48-46 to only 48?
During: maybe this is the main way speciations actually happen in mammals?
"Before: how did 48 become 46 without a tremendous bottleneck?"
ok, i'll connect the dots here since i assumed you would see what i was saying: mongolian wild horses and domestic horses breed and produce fertile offspring despite different chromosome numbers. your assumption that this requires a massive bottleneck is just false.
" If the fork between chimps and humans was before, coincident with, or after the 48->46 fusion seems like a really important trichotomy!"
why? we're more close to chimps than gorillas. you place bit too much emphasize on karyotype
The reason I think it’s important is that either a particular difference in chromosome structures is no significant bar to interbreeding, a major handicap to it, or rules it out entirely. Each of the three hypotheses has problems!
If no significant bar, why did both humans and chimps fixate on different numbers, after the fork?
If it is a major handicap, how did the common ancestor of humans and chimps manage to have such a mix? If it is ruled out entirely, how does a divergence arise without an extreme bottleneck?
If there were just ONE pair of now-mutually-infertile mammalian species differing in this way where what actually happened had been traced and figured out, I would be a lot less puzzled.
I used to think that the 48 vs 46 human chromosome count was a shocking example of scientific groupthink, but I recently looked more carefully at the history, and maybe not quite. My reading is that mainly it was that until the 1950s basically nobody really cared about the number, though indeed one guy in the 1920s reported the human number as 48. It appeared in textbooks, so outsiders assumed it was True, but the (known to insiders) somewhat crappy original estimate was left alone mainly because it didn't seem an important question.
I still don't know the difference between Identity by descent and identity by state. Maybe I'm just dim, and you explained it. Maybe you didn't.
right. basically ibs is simple: does the base or variation match? the provenance is irrelevant. 99.9% of a given human's genome matches another human. why? not because they are closely related in any colloquial sense, but because we're a young species that share common ancestors within the last 300,000 years.
ibd is more recent shared ancestry. go back far enough and it's all the same ancestors, but in the intervening time there has been lots of geographic separation, endogamy, drift and divergence. so traversing the recent pedigrees and genealogy is what looking at 'ibd segments' does. it exposes connections along the last 50 generations or so. really the last 10 in most cases.
You mention that “an example would be the gene that he possesses in common with Europeans: not because of common descent, but because of selection for that gene resulting in separate populations having that gene.” But most likely this gene *has* come down from a common ancestor of him and the Europeans (maybe as long as 4,500 years ago). It *might* have arisen by mutation independently in separate populations, but single origin and common descent therefrom is usually more likely. (Of course, the gene can be selected for in different populations even if it has a single origin in an ancestral population.)
" It *might* have arisen by mutation independently in separate populations, but single origin and common descent therefrom is usually more likely."
you are right about the haplotype, but not necessarily about the SNP. in most cases these concepts arose in the context of a specific marker and not a genomic segment (genomics did not exist)