merge_otu_tables.py duplicate observations in taxonomy

[Émilie Tremblay]

Hello,

I am merging several OTU tables through the merge_otu_tables.py command. 

python /isilon/biodiversity/pipelines/qiime-1.7.0/biom-format-1.1.2-release/bin/convert_biom.py -i /home/CFIA-ACIA/tremblaye/NGS_for_Fungi_Detection/dataAn/merge_OTU/ITS_merged/biom/ITS_MERGED.biom -o $ITS_merged_tsv/ITS_merged.tsv -b --header_key="taxonomy"

but when I look into the table generated, I see that most taxonomy lines have replicates.

When reading the manual, I can see that it should not happen:

To combine multiple BIOM tables into a single BIOM table, you can use merge_otu_tables.py. The main thing that you need to watch out for here is that the OTU ids and sample ids are compatible in each of the tables. If they are overlapping (e.g., you have OTU1 in more than one table), their counts will be summed.

Is there a way that I can tell qime to put them togheter?

I am afraid that it will affect my dowstream analysis (stats).

Thanks

[Émilie Tremblay]

I think that it is because there are more than one sequence with the same name (gi identifiers) but will that affect my analyses?

[Daniel McDonald]

Hi Émilie,

Can you send the output of "biom summarize-table -i /path/to/your/table" for both tables please? 

For 16S analyses, it isn't unusual for multiple OTUs to have the same taxonomy. Is the issue though that you have replicated identifiers?

Best,

Daniel

[Émilie Tremblay]

Hi, I am not using 16S.
Identifiers are not replicated but taxonomy is.
For example, I will see multiple rows with this line in my taxonomy column;

k__Fungi; p__Basidiomycota; c__Agaricomycetes; o__Agaricales; f__Pleurotaceae; g__Pleurotus; s__Pleurotus dryinus

I was able to see duplicates after converting my table back into tsv format and open it in excel and look for replicates.

May you please definte which two tables you want me to post.

My input is a bunch of tables (15 tables) and my output is the merged tables.

Thanks al ot!!

[Émilie Tremblay]

In the meantime here is the summary for my meregd table :)

(see attached, please)

Thanks!!

[Daniel McDonald]

Hi Émilie,

I'm not seeing anything unusual with the output...? For non-16S data, its also not unexpected that multiple OTUs will have the same taxonomy. Can you share with me (direct to my email if you'd like) one of the tables that is problematic?

Best,

Daniel