Using Qiime OTU table as input for metacoder

81 views
Skip to first unread message

ladyp...@gmail.com

unread,
Sep 28, 2017, 5:35:20 PM9/28/17
to metacoder
Hi There!

I ran across the metacoder package and this looks like a really neat way to represent my taxonomic classifications. I was going through the MetacodeR tutorial and I was wondering if this package supports output directly from Qiime. I have parsed my entire OTU table to look like this:



OTU_ID    M40.A    M40.B    M40.C    M41    M80.A    M80.B    M80.C    Root    Phylum    Class    Order    Family    Genus    Species
555371    1    1    0    1    1    0    0    Archaea     Crenarchaeota     Thaumarchaeota     Cenarchaeales     Cenarchaeaceae     Nitrosopumilus     
553746    0    1    0    1    0    1    0    Archaea     Crenarchaeota     Thaumarchaeota     Cenarchaeales     Cenarchaeaceae     Nitrosopumilus     
752517    0    0    0    0    0    0    0    Archaea     Euryarchaeota     Thermoplasmata     E2     Marine group II          
142578    0    0    0    0    0    0    0    Bacteria     Proteobacteria     Deltaproteobacteria     Desulfovibrionales     Desulfovibrionaceae     Desulfovibrio     cavernae
544563    1    0    0    0    0    0    0    Bacteria     Bacteroidetes     Flavobacteriia     Flavobacteriales     Flavobacteriaceae     Robiginitalea     
1040290    1    0    1    0    0    1    0    Bacteria     Chloroflexi     Anaerolineae     CFB-26               
1014140    0    0    0    0    0    2    0    Bacteria     Cyanobacteria     Oscillatoriophycideae     Chroococcales               
342849    0    0    0    0    0    0    0    Bacteria     Cyanobacteria     Synechococcophycideae     Pseudanabaenales     Pseudanabaenaceae     Halomicronema     
1827890    0    2    2    0    0    1    0    Bacteria     Planctomycetes     Phycisphaerae     Phycisphaerales     Phycisphaeraceae          
3950306    0    0    0    0    0    0    0    Bacteria     Proteobacteria     Alphaproteobacteria               
161272    0    0    0    0    1    1    1    Bacteria     Proteobacteria     Alphaproteobacteria     Rhodobacterales     Rhodobacteraceae          
286163    0    0    0    0    2    0    0    Bacteria     Proteobacteria     Alphaproteobacteria     Rhodobacterales     Hyphomonadaceae     Maricaulis     
583673    0    0    0    1    0    0    0    Bacteria     Proteobacteria     Alphaproteobacteria     Rhodobacterales     Rhodobacteraceae     Rhodobacter     
513900    0    0    1    0    0    0    0    Bacteria     Proteobacteria     Alphaproteobacteria     BD7-3               
994959    0    0    0    0    2    0    1    Bacteria     Proteobacteria     Alphaproteobacteria     Rhizobiales     Rhizobiaceae          
567885    0    0    0    0    0    2    0    Bacteria     Proteobacteria     Deltaproteobacteria     Myxococcales               
660200    0    0    0    0    0    0    1    Bacteria     Proteobacteria     Alphaproteobacteria                    
140574    0    0    0    0    0    0    0    Bacteria     Proteobacteria     Alphaproteobacteria     Rhodobacterales     Rhodobacteraceae       


____________
As you can see, not everything goes down to order, genus or species, while some do. Also, will metacoder be able to pull up groups of samples, such as I have above? For example, if I wanted to only look at samples M40.A, M40.B, M40.C can I do this? Or M40.A, M40.B, M40.C only the phylum Cyanobacteria, as another example.

Thank you very much for any insights or help.

Sincerely,
Joany







Zachary Foster

unread,
Sep 29, 2017, 5:36:27 PM9/29/17
to metacoder
Hello Joany,

There is no dedicated parser for that format, but the new `taxa` package has a parser that can easily handle that format. I would recommend using the development versions of `metacoder` and `taxa`. I have attached an RStudio project with an R Markdown that shows how to parse, plot, and subset that data. I also attached the rendered output of the R markdown. Take a look and let me know if you have questions.

-Zach
2017_09_29--joany--parsing.zip
example.pdf
Reply all
Reply to author
Forward
0 new messages