5 ms·

This has been known for a long time. It's a myth that mtDNA only comes from the mother and was based on a misreading of an old paper decades ago. I don't know w

This has been known for a long time. It's a myth that mtDNA only comes from the mother and was based on a misreading of an old paper decades ago. I don't know why the myth has so much traction among geneticists who should know better. Many claims have been made deriving from these assumptions. Nearly all research purporting to date past branches of populations based on mtDNA mutations by assuming a maternal clone is passed, and any differences are mutations, is completely off. All that research is and always has been invalid. 1996 paper on paternal mtDNA inheritance: http://www.pnas.org/content/93/24/13859.full http://www.pnas.org/content/93/24/13859.full 2002 paper on paternal mtDNA inheritance: https://www.newscientist.com/article/dn2716-mitochondria-can-be-inherited-from-both-parents/ https://www.newscientist.com/article/dn2716-mitochondria-can... Wikipedia's been updating this article since 2006: https://en.wikipedia.org/wiki/Paternal_mtDNA_transmission https://en.wikipedia.org/wiki/Paternal_mtDNA_transmission

samatman 8y ago
> Nearly all research purporting to date past branches of populations based on mtDNA mutations by assuming a maternal clone is passed, and any differences are mutations, is completely off. All that research is and always has been invalid. This is a drastic and totally implausible statement that is not supported by any serious geneticist, nor by the Wikipedia article you've linked to. mtDNA does not have to come exclusively, 100%, no-exceptions-ever from the mother, in order for population research which uses to be valid! 99% is quite sufficient. It is at least 99.9% accurate: as the Wikipedia article helpfully points out, there are 1000x as many mitochondria in an egg as in a sperm.
- droithomme 8y ago
  You're attacking a strawman. mtDNA can be used to track lineages. It's the dating of when branches occurred that is off because it assumes all differences are due to mutations and not paternal contributions.
  - whatshisface 8y ago
    That's a good insight - it's not whether or not the rate of paternal inheritance is high, it's whether or not the rate of paternal inheritance is high relative to the rate of mutation. What are those rates for mDNA?
  - samatman 8y ago
    I am not. The rate of mutation is already not a constant, and so such dating is always approximate, there is a margin of error. If you wish to demonstrate that it is now "completely off" and "invalid", try harder. I call shenanigans.
  - gus_massa 8y ago
    I'd have to take a look at the details, but I'm 99% sure that the important part is that the mitochondrial DNA has no crossover, so all the differences are due to mutations and it is a real tree. With normal DNA the crossover add some changes and in particular it mix the changes in one branch of the tree with the changes in the other branch of the tree, so it's not longer a tree. So my guess is that it will not change the dates too much, specially since he dates are calibrated by historical events. With made up numbers, something like "we know that this population reach here 10000 years ago and has 37 changes and that population reach there 20000 years ago and has 69 changes. This other population has 53 changes so after some complicated math, we estimate that they split 15000-1600 years ago. I guess that the size of the population at the bottlenecks will be slightly reduced, because IIUC they are not calibrated using historical data but using some mathematical models about how the size of the population affect the diversity. I don't expect a big change nevertheless.
- rossdavidh 8y ago
  It seems also plausible that the cases where paternal mtDNA are passed on, would be disproportionately likely to have mitochondrial disease and thus eventually die out. We don't know that this is so, but it seems more likely true than not, given what we do know. If so, they would not really impact research done over evolutionary timescales.
JohnJamesRambo 8y ago
How do you know paternal mitochondrial dna transmission isn’t so rare that it has no real measurable effect on population mapping?
- droithomme 8y ago
  In humans between 1 in 1000 and 1 in 10,000 of the mtDNA is copied from the father, and how much and which parts seems to depend a lot on the particular father, perhaps suggesting that there exists an as yet undiscovered checksum scheme, and not random chance. Most mtDNA between humans is the same, and mtDNA is only 16,569 base pairs long. So in most offspring the paternal mtDNA contribution of 1-20 base pairs happens to be the same as the maternal mtDNA and one can not distinguish that it is even present. But sometimes those 1-20 base pairs are different. The fallacious assumption that all changes in mtDNA are due to mutations in the mother is completely off. The differences between mother and child are almost always due to the father's contribution, and not to mutations. This has serious consequences for branch dating based on mutation rates.
  - eggie 8y ago
    > undiscovered checksum scheme I don't think checksum is a fair metaphor for anything happening in biological systems. > Most mtDNA between humans is the same And most DNA is also the same. A percent difference is the kind of thing that would require millions of years of divergence to develop. > mtDNA is only 16,569 base pairs long This is really not so precise. There will be a mixture of lengths due to insertions and deletions, even in the same individual, even in the same cell. > The differences between mother and child are almost always due to the father's contribution, and not to mutations. Do you have a citation for this?
    - Retric 8y ago
      Checksum is a solid analogy, take natural abortion. If a fetus fails specific challenges it’s removed. This is very common, with a significant portion of pregnancy’s ending this way.
    - 8y ago

samatman 8y ago

> Nearly all research purporting to date past branches of populations based on mtDNA mutations by assuming a maternal clone is passed, and any differences are mutations, is completely off. All that research is and always has been invalid. This is a drastic and totally implausible statement that is not supported by any serious geneticist, nor by the Wikipedia article you've linked to. mtDNA does not have to come exclusively, 100%, no-exceptions-ever from the mother, in order for population research which uses to be valid! 99% is quite sufficient. It is at least 99.9% accurate: as the Wikipedia article helpfully points out, there are 1000x as many mitochondria in an egg as in a sperm.

droithomme 8y ago
You're attacking a strawman. mtDNA can be used to track lineages. It's the dating of when branches occurred that is off because it assumes all differences are due to mutations and not paternal contributions.
- whatshisface 8y ago
  That's a good insight - it's not whether or not the rate of paternal inheritance is high, it's whether or not the rate of paternal inheritance is high relative to the rate of mutation. What are those rates for mDNA?
- samatman 8y ago
  I am not. The rate of mutation is already not a constant, and so such dating is always approximate, there is a margin of error. If you wish to demonstrate that it is now "completely off" and "invalid", try harder. I call shenanigans.
- gus_massa 8y ago
  I'd have to take a look at the details, but I'm 99% sure that the important part is that the mitochondrial DNA has no crossover, so all the differences are due to mutations and it is a real tree. With normal DNA the crossover add some changes and in particular it mix the changes in one branch of the tree with the changes in the other branch of the tree, so it's not longer a tree. So my guess is that it will not change the dates too much, specially since he dates are calibrated by historical events. With made up numbers, something like "we know that this population reach here 10000 years ago and has 37 changes and that population reach there 20000 years ago and has 69 changes. This other population has 53 changes so after some complicated math, we estimate that they split 15000-1600 years ago. I guess that the size of the population at the bottlenecks will be slightly reduced, because IIUC they are not calibrated using historical data but using some mathematical models about how the size of the population affect the diversity. I don't expect a big change nevertheless.
rossdavidh 8y ago
It seems also plausible that the cases where paternal mtDNA are passed on, would be disproportionately likely to have mitochondrial disease and thus eventually die out. We don't know that this is so, but it seems more likely true than not, given what we do know. If so, they would not really impact research done over evolutionary timescales.

JohnJamesRambo 8y ago

How do you know paternal mitochondrial dna transmission isn’t so rare that it has no real measurable effect on population mapping?

droithomme 8y ago
In humans between 1 in 1000 and 1 in 10,000 of the mtDNA is copied from the father, and how much and which parts seems to depend a lot on the particular father, perhaps suggesting that there exists an as yet undiscovered checksum scheme, and not random chance. Most mtDNA between humans is the same, and mtDNA is only 16,569 base pairs long. So in most offspring the paternal mtDNA contribution of 1-20 base pairs happens to be the same as the maternal mtDNA and one can not distinguish that it is even present. But sometimes those 1-20 base pairs are different. The fallacious assumption that all changes in mtDNA are due to mutations in the mother is completely off. The differences between mother and child are almost always due to the father's contribution, and not to mutations. This has serious consequences for branch dating based on mutation rates.
- eggie 8y ago
  > undiscovered checksum scheme I don't think checksum is a fair metaphor for anything happening in biological systems. > Most mtDNA between humans is the same And most DNA is also the same. A percent difference is the kind of thing that would require millions of years of divergence to develop. > mtDNA is only 16,569 base pairs long This is really not so precise. There will be a mixture of lengths due to insertions and deletions, even in the same individual, even in the same cell. > The differences between mother and child are almost always due to the father's contribution, and not to mutations. Do you have a citation for this?
  - Retric 8y ago
    Checksum is a solid analogy, take natural abortion. If a fetus fails specific challenges it’s removed. This is very common, with a significant portion of pregnancy’s ending this way.
  - 8y ago