Bacterial phylogeny download for a list of species

Question

[susan.klein]

Hi,

when analysing 16s metagenomic data, its easy to align all of your otu sequences and build a tree in order to e.g. calculate UniFrac distances. However, with shotgun metagenomic results I cannot find a simple way to get a tree of the taxa I am interested in. I've looked at the usual databases, Silva, NCBI, RDP etc. but none of them seem to have a facility to obtain a tree from a list of taxa. Does anyone know a way to do this? I could download all the 16s for my list of taxa and align and draw a tree from that, but hoping there is a way to do it from the list of taxa alone.

Thanks for any help,

T.

[jhc]

for NCBI trees, check ete-ncbiquery

Answer 1

[Andrzej Zielezinski]

NCBI Taxonomy database provides a feature named Common Tree, where you can provide a list of taxonomy identifiers to create a tree of species you are interested in. The tree will be displayed in your browser and you can save it to a file (in text or Newick format).

For example, here is a tree of human, mouse and chimp:

Euarchontoglires
+ Glires
+ + Mus musculus
+ Primates
+ + Homininae
+ + + Homo
+ + + + Homo sapiens
+ + + Pan
+ + + + Pan troglodytes
-----------------------------------

Newick format:

(
'Mus musculus':4,
(
'Homo sapiens':4,
'Pan troglodytes':4
)Homininae:4
)Euarchontoglires:4;

Answer 2

[susan.klein]

Hi,

yes I've seen that.. these trees have no branch lengths so can't be used to calculate UniFrac. There is also the added complication of obtaining the 'taxonomy ID' which is of course different fromt he GG ID.

I'm working on a script to solve this atm.

Thanks,

S

Answer 3

[susan.klein]

Found a script that do this, someone might find it useful. I outputs a fasta file of 16s sequences matching (or close as possible matching) the input taxonomy. The fastas can then be aligned and a tree produced.

S

the link died.. here is another link to the code.

https://github.com/theo-allnutt-bioinformatics/scripts/blob/master/tax2_16s.py