An Exercise in Similarity

Post of the Month: August 2005

Subject:    | Re: Request
Date:       | 4 August 2005
Message-ID: | NailedIt@demKBykoL

John Harshman wrote:

> > I will post the sequences in this thread, five posts, one set of
> > nucleotides per post. I'll leave the subject line untouched. The
> > first line of each post should read SAMPLE A1, SAMPLE A2, ....
> > SAMPLE A5. A blank line will follow, then the nucleotides, there
> > should be nothing else in the posts. I've also trimmed the numbering
> > on the left side, I was afraid it was getting close to wrapping in
> > some news readers. There is one each of
> >
> > Homo sapiens
> > Pan troglodytes
> > Gorilla gorilla
> > Pongo pygmaeus
> > Hylobates klossii
> >
> > There is no significance to the sequential scheme A1,A2,A3,A4,A5 and
> > I've intentionally mixed them up so there would not be.
> >
> OK, here are my identifications:
>
> A3 = Hylobates

Correct.

> A5 = Pongo

Correct.

> A1 = Gorilla

Correct.

> A2, A4 = Homo and Pan

A2 is Homo, A4 is Pan.

I also had Pan paniscus, I think you would have lock down two Pan representatives with absolute certainty if I had added it to the mix, but of course there would have been no way to distinguish between Pan troglodytes and Pan paniscus.

Thanks for the effort, I find this to have been a noteworthy exercise.

> Like I said, the first two are easy, the third less clear. And the last
> two are interchangeable; there's no possible way to tell them apart just
> by the shape of the tree.
>
> Here's how I did this:
>
> First I had to massage slightly to put it into analyzable form using the
> program PAUP. I put all the sequences in one text file. I put a _
> between SAMPLE and A1, etc., because PAUP recognizes names as everything
> before the first space. I added this to the beginning of the file:
>
> #NEXUS
>
> BEGIN DATA;
> DIMENSIONS NTAX=5 NCHAR=1002;
> FORMAT DATATYPE=DNA ;
>
> MATRIX
>
> The first line reassures PAUP that the file is in its standard format.
> The next lines identify a NEXUS data block, tell the program the number
> of species and characters, and inform it that the data are DNA. The last
> line tells it that the actual data are starting.
>
> Then I added this to the end:
>
> ;
> end;
>
> Those two lines tell the program that the data are ending, and that the
> DATA block is ending.
>
> Then I executed the file in PAUP, which means that I loaded the data
> into the program memory, and then did one of the simplest analyses,
> neighbor joining. Other methods gave me the same tree. Here's the tree:
>
> Neighbor-joining tree:

> /------- SAMPLE A1
> |    /--------------------------- SAMPLE A2
> | /--+------------- SAMPLE A4
> \-+              /--------------------[...]------------------- SAMPLE A3
>   \--------------+------------------------- SAMPLE A5

> I had to shorten the branch to A3 so it would fit on this page; that's
> what the [...] means. OK, I pick the longest branch, A3 as Hylobates.
> Strictly speaking, this only works if there's a uniform rate of
> evolution, which is a bad assumption. But usually it will be close
> enough to make me confident in my pick. Of course, we will usually have
> a designated outgroup in the data, so I wouldn't have to make such a
> guess. All else follows from that. The first branch from A3 leads to A5,
> so that's Pongo. The next branch leads to A1, so that is suggested to be
> Gorilla. Note, however, that the branch separating A1 from A2-A4 is very
> short, so it's not very reliable. If A1 is Gorilla, then A2 and A4 are
> Homo and Pan, though we can't say which is which.
>
> Anyway, I had only one choice to make, and that was which one is
> Hylobates. All else follows from the shape of the tree. Picking
> Hylobates requires the single additional assumption that the rate of
> evolution doesn't vary too extremely from one species to another.
>
> So, are these more mitochondrial sequences?

No, a taste receptor gene. Take note that the Homo entry says "candidate taste receptor TAS2R38" all the others say "taste receptor type 2 member 38".

Homo sapiens
http://www.ncbi.nlm.nih.gov/entrez/viewer.fcgi?db=nucleotide&val=AF494231
http://tinyurl.com/6s48vns

Gorilla gorilla
http://www.ncbi.nlm.nih.gov/entrez/viewer.fcgi?db=nucleotide&val=45549380
http://tinyurl.com/adc6p

Hylobates klossii
http://www.ncbi.nlm.nih.gov/entrez/viewer.fcgi?db=nucleotide&val=45549384
http://tinyurl.com/drecx

Pongo pygmaeus
http://www.ncbi.nlm.nih.gov/entrez/viewer.fcgi?db=nucleotide&val=45549382
http://tinyurl.com/9bnpu

Pan troglodytes
http://www.ncbi.nlm.nih.gov/entrez/viewer.fcgi?db=nucleotide&val=45549378
http://tinyurl.com/7nxco

Pan paniscus
http://www.ncbi.nlm.nih.gov/entrez/viewer.fcgi?db=nucleotide&val=45549376
http://tinyurl.com/a6wzw

[Return to the 2005 Posts of the Month]