Naked Science Forum

Life Sciences => Cells, Microbes & Viruses => Topic started by: thedoc on 01/12/2014 19:30:01

Title: Can we run regression algorythms on DNA to find the common ancestor?
Post by: thedoc on 01/12/2014 19:30:01

Hudson Ansley  asked the Naked Scientists:
   
I have had a question for some time that seems like might be interesting for your show to investigate: Would it be possible, perhaps with some future technology, to derive the DNA sequence of a common ancestor given the DNA of two separate species. Naively, it seems like it would, with some algorithm that would use the the similarities and differences, maybe from multiple individuals of each species, to derive the common ancestor sequence. I guess my question is more is there some fundamental reason why this would not be possible.

Given the possibilities of "bringing back" extinct species recently in the news, this would have some interesting implications if it actually could work.
 
I recently found your podcast (http://www.thenakedscientists.com/HTML/podcasts/) and have been both entertained and informed by it; thank you!

Best Regards,
Hudson


   

What do you think?

Title: Re: Can we run regression algorythms on DNA to find the common ancestor?
Post by: RD on 01/12/2014 22:46:19

Quote from: thedoc on 01/12/2014 19:30:01

... derive the DNA sequence of a common ancestor given the DNA of two separate species ... derive the common ancestor sequence. I guess my question is more is there some fundamental reason why this would not be possible.

Only the DNA sequences which have survived to the present day , some would exist in the ancestor which haven't made it.

http://en.wikipedia.org/wiki/Most_recent_common_ancestor#MRCA_of_different_species

Quote from: thedoc on 01/12/2014 19:30:01

...  some algorithm that would use the the similarities and differences, maybe from multiple individuals of each species ...

http://en.wikipedia.org/wiki/Coalescent_theory#Software

Title: Re: Can we run regression algorythms on DNA to find the common ancestor?
Post by: CliffordK on 01/12/2014 23:24:11

Interesting thought. 

If you have just two samples, say human and ape, both of them have evolved over the last 5 to 10 million years, then anything that is identical between the two likely was in the common ancestor.  However, you wouldn't know if differences were due to evolution of the Ape, Human, or both, so you would be left with a number of holes. 

Now, we have many more primates than hominids. 

So, assuming that apes separated from monkeys before hominids separated from apes. 

Then, any identical DNA found in both Apes and Monkeys would also have to be in the last common ancestor for humans.  How complete is the Neanderthal genome?  That would add additional points.

Similar DNA in the Great Apes (Chimps, Orangutans, Gorillas,  Bonobos) would help.

Looking at the Evolutionary Tree
(https://www.thenakedscientists.com/forum/proxy.php?request=http%3A%2F%2Fupload.wikimedia.org%2Fwikipedia%2Fcommons%2Fthumb%2F7%2F7f%2FHominidae_chart.svg%2F400px-Hominidae_chart.svg.png&hash=476ec389fa50d5bce09d583e37507dc9) (http://en.wikipedia.org/wiki/Hominidae)

Chimps (Pan) <--> Gorillas (Gorillini) commonalities gives one the DNA that a common ancestor for humans MUST have (with a few exceptions as two species could receive identical mutations later during the evolution).
Chimps (Pan) <--> Bonobos (Pan) - Assumed separation, about 1.5 million  years ago.
Humans <--> Neanderthals   - Assumed separation, about 500,000 years ago.

Anyway, the more data points the better, both pre and post separation of the species.  As RD mentioned, one wouldn't ever be certain about deletions that BOTH subspecies received, or mutations that both received.