Is it possible to know which otus gets clubbed in one phylotype during summarize_taxa.py

[Rittik Deb]

Hi,

We want to know which of the OTUs get clubbed to one phylotype during summarization step. I searched but their are no log files. I also tried running biom.summary, but it only says how many otus were clubbed, not which. Is it possible to get the info.
The mainreason behind this is we want to run picrust only on dominant phylotypes which only accepts pre summarization biom file. The dominant phylotype choice can only be done post summarization. So it is essential to know this info.

With warm regards

Rittik

[Antonio González Peña]

Which script are you using to group by phylotype? summarize_taxa.py? If that's the case and you know which taxa you want to keep, you could use filter_taxa_from_otu_table.py to get what you want. 

[Rittik Deb]

Hi, 

Thanks for your reply but i guess my question was not clear. I am looking for dominant taxa post summarization. Then I need to find only those otus that contributed to one dominant phylotype. So i need to know which otus are collated for any of the phylotypes during summarization. Is it possible to get that info....

For eg. Which will say x,y,z otus were clubbed to form phylotype j

I am using summarize_taxa.py script.

Thanks in advance

[Antonio González Peña]

There is no direct way to do this but the commands I sent you should help you get that info. Let's imagine that we are interested in p__Firmicutes. The commands should be something like:

filter_taxa_from_otu_table.py -i otu_table.biom -o p__Firmicutes.biom -p p__Firmicutes

biom summarize-table --observations -i p__Firmicutes.biom

Hope this helps. 

[Rittik Deb]

Hi,

Thanks again. But the issue remains that often multiple levels of taxonomic OTUs can get clubbed into a single phylotype thus generating classifications like --g--s. Here it is pretty hard  to filter the otus without knowing which one.
For e.g. in our case after summarization, we get
Enterobacteriaceae__g__s    and   Enterobacteriaceae__Serratia__s
Now these two will have different otus clubbed under them and filtering doesnot help.

I feel slightly surprised to know that the info of which otus gets clubbed under one phylotype is not available or cannot be extracted.

Any other suggestion will be greatly appreciated.

Regards

Rittik

[jonsan]

Hi Rittik,

Maybe what you're looking for is summarization to a more inclusive taxonomic level? For example, the case you give above looks like you've summarized to the species level (level 7), so that Enterobacteriaceae__g__s    and   Enterobacteriaceae__Serratia__s end up as separate lines on the summarized OTU table. But what I infer you're wanting to do is to get everything assigned to the family Enterobacteriaceae rather that to more specific genera and species. To do that you'd simply filter your starting OTU table at a different level. Per Antonio's code:

 filter_taxa_from_otu_table.py -i otu_table.biom -o f__Enterobacteriaceae.biom -p  f__Enterobacteriaceae

This will give you all OTUs classified as Enterobacteriaceae, including those under Enterobacteriaceae__g__s    and   Enterobacteriaceae__Serratia__s

does this make sense?

-jon

[Rittik Deb]

Thanks to both of you for your replies.....

I was under a silly impression that the collation happens based on similarity between OTUs irrespective of taxonomic assignment and hence the final summary taxonomic ID may change based on how many of which otu taxa gets collated....
I figured that is not at all the case, just the same naming (taxonomic) gets collated during summary.....

Sorry for bothering you so many times with my wrong assumptions.

My bad.....hope this will not deter you from responding later.

Rittik

[jonsan]

No worries, and glad it's resolved for you!

-j