Assign Taxonomy step

[Alexandra Puértolas]

Dear Qiime community,

Finally, I reached the last step of my data analysis! I am performing a metagenomic study using four different primer pairs, and I have a little issue on the taxonomy result.

My question is about assign taxonomy to the representative OTU sequences. When I run the assing_taxonomy.py scrip with the blast option, for some of the samples I get the correct taxonomy assignment to the correspondent sequence of my database, but for some samples (sometimes are quite a lot of them), I don't get a taxonomy-sequence result, I just get as a result the name of the fasta database I used on this step with a number taht sometimes is the same but different in other otus (I have created my own database, which is a fasta file of a onw-curated collection of unique sequences. Not all the sequences on the reference database have the same lenght of my Illumina reads - for example, on the database I have sequences for the entire ITS region but I have only sequenced the ITS1 spacer).

I am a bit confused about what does it mean... I don't know if that is because the consensus sequence of this otu fits to more than one sequence of the database at the same time, and it cannot return a specific result.

Here is what I am talking about (see the otus 'denovo189, denovo370 and denovo430). Any idea of what does is it?

denovo748  Eukaryota  Stramenopiles Oomycetes PeronosporalesPhytophthora PHYTOPHTHORA_TENTACULATA_CBS_41296_ITS 9 E-024 PHYTOPHTHORA_TENTACULATA_CBS_41296_ITS

denovo189         None 4 E-011 786_DBsingles.fasta

denovo370   None  0.000000003 786_DBsingles.fasta

denovo430  None 4 E-135 438_DBsingles.fasta

denovo438 Eukaryota Stramenopiles Oomycetes Peronosporales Phytophthora PD_00420_ITS_Phytophthora_taxon_Pgchlamydo 4E-051 PD_00420_ITS_Phytophthora_taxon_Pgchlamydo

Thanks a lot in advance!

Alexandra.

[Alexandra Puértolas]

I have also tried the uclust method in assign_taxonomy.py, and here I don't get that problem with the assigment of the result to the otus. The problem here is that on the log file of assign_taxonomy I read the complete taxonomy but I don't see it in the rep_seq_tax_assigned.txt file, and in consequence, it doesn't appear when I merge the otu table with the taxonomy assignment.

Does anyone know why is that? 

I have attached the assign_taxonomy file and log file, as well as the otu table.

Thank you!

[zech xu]

Hi Alexandra,

What reference ITS database are you using? could you grab a representative sequence of the OTUs that get weird or no taxonomy assignment and blast it against NCBI to see if it truely matches ITS?

[Alexandra Puértolas]

Hi Zech Xu

Thank you for your reply. I have checked and these sequences labelled on this way belong to the ITS region but they look like they are not my target organisms. The database I am using is a complilation of sequences I downloaded from different databases and papers for the four loci I have studied.

I have been checking and I will use the uclust method to assign the taxonomy. I am loosign diversity when I use the blast method instead of uclust, and with it I can change the parameters, and modify the similarity to 0.8 (with blast I cannot do that and it will be always 0.9, and I think that this is why I have more diversity with uclust).

Do you know how can I fix the step to make an otu table with the taxonomy without loosing the name of each reference sequence?

Thanks a lot for your help!

Alexandra

[zech xu]

Hi Alexandra,

To get the ref seq ID that your ITS match, you can get this info in the raw format output from uclust. If you look at the ouput dir from assign_taxonomy.py command, there is a log file. You have to parse that file to get which rep seq matches which ref seq.

[Alexandra Puértolas]

Hi! Thank you for the information!

I saw the log file with the results of the taxonomy. How can I parse this information and merge it in an otu table to get in there the otu counts and the species assignments? Is there a Qiime or other script to do it?

Thanks a lot for your help

Alexandra