How Reliable is Ancient DNA?

Alex266

Junior Member
Messages
5
Reaction score
2
Points
3
DNA decays over time. A Harvard trained geneticist had this to say about DNA decay:

“In the lab, even if we stored DNA in a −20° C manual defrost (i.e., non-frost-free) freezer, we would discard DNA samples that were over a year old. Beyond a year, the DNA samples gave unreliable and unpredictable results.”


How is this problem addressed in Paleogenetics? How do we know that ancient DNA results are accurate?
 
There is quality control already at the point of data procession: https://academic.oup.com/bib/article/15/6/879/180439?login=false

The worst thing to use is Imputation to fill the gaps (Missing alleles), because modern populations cannot be compared to ancient ones in their specific SNPs. For example ancient individuals can have a different phenotype as the modern populations they are related to and imputation can create a false picture of those people and modernize them.

But there are some other criteria that can be used as a proof of quality for ancient DNA. For example similarity. When you take 5 samples from a neolithic tomb from 5 different skulls and they cluster in a PCA or are very similar in their traits, it is likely that their DNA is relatively intact. The errors in the DNA will likely not occur at the same places in the DNA strain.
There is the possibility of outliers from other populations or mixed individuals, but it is unlikely that you would find 5 individuals with completely different ancestry in a neolithic tomb.

An example of higher, medium and bad quality Neanderthal samples (Mezmaiskaya, Forbes Quarry, Belgium):



Using bad quality samples enhances the likelihood of erroneous matches.
 
Besides this coherence Doggerland refers to, some SNPs may get decent mapping quality, whereas others may not. So I suppose they would be filtered also based on the quality of the alignment, which must be minimal. Despite possible errors, the results as a whole should be reliable, if the coverage is enough.

In the case of uniparental markers, they would be evidenced by a set of results that overall must respect the known phylogeny, not for isolated positive/negative results.

Another common problem involving ancient DNA is the deamination, which causes apparent C>T and G>A mutations. But deaminated samples are usually evident, i.e., heavy in these apparent mutations, and the events are mostly one read and/or occur at the end of the read.
 
Bear with me for a moment, as strange as it sounds, can we rule out the fact that environmental factors might cause samples in the same geographical area to decay in the same manner?

Not my argument but an argument that I’ve heard.
 
This is an interesting article about the problem and you are not so wrong:

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3685887/

Maybe the same microbe DNA at the site or climate zone is contaminating DNA in the same way.
And the burial practices could also be responsible for the way the DNA is altered? Who knows.

This is fatal for many ancestry markers that are used by hobbyist calculators and also predictions about phenotype. A transition from G to A or C to T can easily make a brown to blue eyes or a light skin to dark or the other way around.
For example if all the SNP alleles for skin color would be damaged like described in the article, the person would be some what intermediate/brown skin color, like it is the case with many ancient European samples.
Red hair is also often caused by an A alelle in a SNP, so maybe some red haired samples are in fact not red haired.

If this is the norm for ancient DNA, we ware all just victims of a playful illusion ;)

Also interesting: "Our analyses suggest that the Actinobacteria, and especially members of the Streptomycetales, contribute the majority of sequences in the DNA extracted from the Neanderthal fossil Vi33.16."

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3643900/
 
This is an interesting article about the problem and you are not so wrong:

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3685887/

Maybe the same microbe DNA at the site or climate zone is contaminating DNA in the same way.
And the burial practices could also be responsible for the way the DNA is altered? Who knows.

This is fatal for many ancestry markers that are used by hobbyist calculators and also predictions about phenotype. A transition from G to A or C to T can easily make a brown to blue eyes or a light skin to dark or the other way around.
For example if all the SNP alleles for skin color would be damaged like described in the article, the person would be some what intermediate/brown skin color, like it is the case with many ancient European samples.
Red hair is also often caused by an A alelle in a SNP, so maybe some red haired samples are in fact not red haired.

If this is the norm for ancient DNA, we ware all just victims of a playful illusion ;)

Also interesting: "Our analyses suggest that the Actinobacteria, and especially members of the Streptomycetales, contribute the majority of sequences in the DNA extracted from the Neanderthal fossil Vi33.16."

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3643900/

Doggerland: Interesting article. A couple of Caveats are in order in my view: 1) This paper was from 2013. 2) DNA research methods and technology have improved since then, 3) DNA has been sequenced in regions that 9-10 years ago none had as of that time been sequenced (e.g., ancient genomes from sub-Saharan Africa have been published since that time).

So I think lots of ancient genomes that were published 9-10 years ago and easily be re-examined with newer technologies as they come on line and results can be re-calibrated, etc.
 
Most if not all of the ancient DNA are from inside bones or from teeth so it's sealed and not exposed to air. So while there is some degradation it is not as bad as it could be.
 

This thread has been viewed 2080 times.

Back
Top