Pick_open_reference and OTUs

[Deb Colman]

Hi everyone! I have some doubts about the results of pick_open_reference and pick_de_novo. After I ran these scripts I have 1 otu_table, when I checked the folders, in uclust_picked_otus (for example), otu´s are represented for a mix of sequences, in my case, from different lagoons. Shouldn´t the otu´s have sequences of a determinated sample?

I attach you the file.

 Also, I don´t know if QIIME can distinguish otu´s from different samples, if I work on 1 unique file (seqs.fna, with all sequences from the 7 lagoons).

Respect to pick_open_reference, I don´t clear the difference between otu_table_mc2 and otu_table_mc2_w_tax and which I can use  to summarize_taxa_through_plots.py

Is it ok if the pick_open_reference showed me 4 folders: step_1_otus, step_2_otus, step_3_otus and step_4_otus? In other analysis, I just had 2 folder, the numer 1 and number 4.

For now, that´s all! Thanks a lot!! 

[justink]

On Thursday, June 15, 2017 at 7:21:26 AM UTC-5, Deb Colman wrote:

Hi everyone! I have some doubts about the results of pick_open_reference and pick_de_novo. After I ran these scripts I have 1 otu_table, when I checked the folders, in uclust_picked_otus (for example), otu´s are represented for a mix of sequences, in my case, from different lagoons. Shouldn´t the otu´s have sequences of a determinated sample?

I attach you the file.

That's a good point, and it means that your analysis of a sample depends on what other samples you analyze with it. Super gross, but in my experience not often a major effect on the results you end up writing about in papers or making figures of. I think there's literature on this too, though I don't have any citations off the top of my head. If you don't like it, closed reference otu picking may appeal to you. It's hard to get away entirely from common OTUs though, because it's nice to be able to e.g. say that sample 1 has 100 of OTU1, and sample 2 has 150 of the same OTU 1.

 

 Also, I don´t know if QIIME can distinguish otu´s from different samples, if I work on 1 unique file (seqs.fna, with all sequences from the 7 lagoons).

seqs.fna probably has the sample information in the header of each sequence, e.g.: >sample1_100 ... 

Respect to pick_open_reference, I don´t clear the difference between otu_table_mc2 and otu_table_mc2_w_tax and which I can use  to summarize_taxa_through_plots.py

Is it ok if the pick_open_reference showed me 4 folders: step_1_otus, step_2_otus, step_3_otus and step_4_otus? In other analysis, I just had 2 folder, the numer 1 and number 4.

http://qiime.org/scripts/pick_open_reference_otus.html has some info, check out step 6. Basically though, w_tax means with taxonomy—use that for  summarize_taxa_through_plots.py