DNA similarity to humans

[Peterski]

And when it comes to differences between humans, according to prof. Luiza Handschuh:

"(...) The genomes of humans, animals, plants, and bacteria differ in length and nucleotide sequence, but the more closely related the species, the more of similar fragments they have. The human genome is large and contains over 3 billion nucleotides. If we compare the genomes of individual people who are not close relatives, we will find that they are identical on average in 99.85%. However, the seemingly small percentage of differences (0.15%) related to the occurrence of various types of mutations or variable (polymorphic) sites, amounts to as much as 4.5 million nucleotides. Many of these differences do not translate in any way into the physical characteristics of the organism, but they are valuable from the point of view of genetics and population genomics, forensic medicine, and genetic genealogy. (...)"

Two humans are identical on average in 99.85%, while I think the minimum is 99.5%.

[Peterski]

Do you think that the video is really accurate?

I'm a bit surprised for example that the Koala ranked the same as the Chicken.

Shouldn't all mammals be closer than any bird?

Edit:

Well actually Koala is the last mammal in that list and Chicken is the first bird.

[Nurzat]

if dandelion has x% similarity to humans, measured at the amount of its own DNA, then that will be an infinitesimal amount if measured at what % is that of human DNA. so even if say some small thing has 10% of its genes that are also in humans, humans won't have 10% of their genes that are also in the small thing, as their (the humans') genome will be huge in comparison.

simplified practical example:

being #1 has 5 genes and 2 are common to being #2 (i.e. 40% of its genome)
being #2 has 200 genes and those 2 are common to being #1 (i.e. 1% of its genome)

[AndreiDNA]

And when it comes to differences between humans, according to prof. Luiza Handschuh:

"(...) The genomes of humans, animals, plants, and bacteria differ in length and nucleotide sequence, but the more closely related the species, the more of similar fragments they have. The human genome is large and contains over 3 billion nucleotides. If we compare the genomes of individual people who are not close relatives, we will find that they are identical on average in 99.85%. However, the seemingly small percentage of differences (0.15%) related to the occurrence of various types of mutations or variable (polymorphic) sites, amounts to as much as 4.5 million nucleotides. Many of these differences do not translate in any way into the physical characteristics of the organism, but they are valuable from the point of view of genetics and population genomics, forensic medicine, and genetic genealogy. (...)"

Two humans are identical on average in 99.85%, while I think the minimum is 99.5%.

They have a very peculiar way of determening the similarity. They obviously don't use a method that compares genotypes, because then there is no way humans can be 99.85% identical, the number would be waay waay lower. Perhaps they base this number on DNA structure, counting areas where there are matching polymorphisms as "similar areas". But even if that's the case, there is no way creatures like octopi would be 50% similar genetically to humans, this is still way too high - their genomes aren't structurally similar to ours at all. So really what I'm thinking they are counting is the total ratios of DNA proteins found in humans vs various other species. Take videos and statements like this with a grain of salt. F2 of Africans to Europeans is 0.2, meaning that across the SNPs selected for the dataset, on average the genotype similarity between Africans and Europeans is 55% (1-sqrt(0.2))

[Rogan]

Of course not. Classic pop science trick, oversimplify until it sounds impressive enough to sell.

Humans and bacteria share some genes (like for basic cell metabolism, DNA replication, etc, etc.), but the majority of their DNA is not homologous, it can't be aligned at all. If you only compare shared genes, the number looks high, but if you compare entire genomes, the number plummets to nearly 0% cuz bacteria don't have most of the genomic elements humans do (introns, regulatory regions, etc.). The denominator matters.

When they say "humans share 50% with bananas," that doesn't mean half your DNA looks like a banana's. It means half of the types of genes are found in both species.The DNA sequences themselves are often very different, just the functions are conserved.

[Rogan]

Also, at large evolutionary distances DNA changes so much that only a few fundamental genes can still be aligned. This makes simple % comparisons meaningless, beyond a certain point (e.g., bacteria vs. human), there is simply no continuous alignment possible.

[Peterski]

F2 of Africans to Europeans is 0.2, meaning that across the SNPs selected for the dataset, on average the genotype similarity between Africans and Europeans is 55% (1-sqrt(0.2))

Yes but this is only across the SNPs selected for the dataset which is only 1,240,000 SNPs if you are using Reich's AADR dataset.

[Peterski]

Of course not. Classic pop science trick, oversimplify until it sounds impressive enough to sell.

Humans and bacteria share some genes (like for basic cell metabolism, DNA replication, etc, etc.), but the majority of their DNA is not homologous, it can't be aligned at all. If you only compare shared genes, the number looks high, but if you compare entire genomes, the number plummets to nearly 0% cuz bacteria don't have most of the genomic elements humans do (introns, regulatory regions, etc.). The denominator matters.

When they say "humans share 50% with bananas," that doesn't mean half your DNA looks like a banana's. It means half of the types of genes are found in both species.The DNA sequences themselves are often very different, just the functions are conserved.

Also, at large evolutionary distances DNA changes so much that only a few fundamental genes can still be aligned. This makes simple % comparisons meaningless, beyond a certain point (e.g., bacteria vs. human), there is simply no continuous alignment possible.

Okay, but for animals such as apes, monkeys and other mammals their numbers are correct in your opinion?

[Rogan]

Okay, but for animals such as apes, monkeys and other mammals their numbers are correct in your opinion?

Yes for apes and other mammals the numbers are roughly correct, although still oversimplified. Mice for instance do share about 85% of their coding genes with us, which is why they're used in human biomedical studies, but that doesn't mean 85% of our total DNA is the same. We don't share 85% of total genom sequence.

For non-mammals those % are largely meaningless because too much of the DNA isn't directly comparable. Pop-sci mags tend to ignore which parts of the genome they're comparing and often mix coding and noncoding regions.

We definitely don't share 35–50% of our genes with bananas or some rice lol, when only the core cellular machinery is comparable and overlaps (stuff every living cell needs to survive), and that makes such comparisons almost meaningless.
Saying we overlap by 35–50% in those universal genes is technically true, but it tells you nothing about overall DNA similarity or evolutionary proximity.