I am trying to perform picrust analysis on nephele results that were given to me by a past colleague. My colleague provided me with a biom table (otu_table_mc2_w_tax_no_pynast_failures.biom), map.txt and tree.tre. That’s it.
I am trying to massage these files into a format I can use with picrust2.
I attempted to import these files into qiime2 and I was able to extract the feature table and taxonomy from the biom file. However, the qiime2-picrust2 pipeline appears to want ASV count and ASV seq.
I saw another article that said the issue was my otu_table_mc2_w_tax_no_pynast_failures.biom was in a newer .biom format so I converted it to json and that still didn’t work.
I’ve already imported these nephele results into R as a phyloseq object so I began looking for options to create a picrust2 OTU table from a phyloseq object - I came up empty-handed.
I tried to install QIIME1 but given that I am a mac user (and it is wayyyyy outdated) that was not a viable feat.
Is there ANY way for me to get a picrust2 compatible OTU table from nephele results or a phyloseq object (created from the nephele results)
Thanks in advance for ANY help you can offer. I’ve been wracking my brain for days now and I am about to throw my computer out the window. I apologize in advance if this question has been asked before - I did look!
I think you might be out of luck without the sequences of each OTU/ASV. From the Key Limitations of picrust:
Predictions are limited by how your study sequences can be placed in the reference tree. By default EPA-NG is used to place study sequences, which requires several considerations to be taken into account. ...
While it might be possible to use the tree.tre file you already have, the pipeline is designed to use the sequences of your features.
Is there any chance your former colleague has the old OTU sequences hiding somewhere?
Or perhaps they have the raw data so that you could reprocess into a brand new ASV table?
Darn - I was afraid of that. This colleague was able to perform picrust analysis on different tissues within the same animals so I have reached out to him to find out what key piece of information he is withholding (kidding!)
Thanks so much for taking the time out to respond.
I was able to get some more information from my colleague (yay!) But I am still having trouble. I am not familiar with QIIME1 at all… I’ve tried going through the old tutorials and walk-throughs, but I am still struggling!
My coworker gave me access to a google drive containing the following:
Step 1 OTU’s
Abundance_sorted.fna
Abundance_sorted.uc
Failures.fasta
Ref_clustered.uc
Step1_rep_set.fna
Step 2 OTU’s
Denovo_abundace_sorted.fna
Denovo_abundance_sorted.uc
Denovo_smallmem_clustered.uc
Step4_rep_set.fna
Pynast_Aligned_Seqs
rep_set_aligned_pfiltered.fasta
rep_set_aligned.fasta
rep_set_failures.fasta
Other Files
new_refseqs.fna
otu_metadata_added_semecolon.biom
otu_metadata_added.biom
otu_table_mc2_w_tax_no_pynast_failures.biom
otu_table_mc2_w_tax.biom
otu_table_mc2.biom
rep_set.tre
table.from_txt_json.biom
phew
So from what I can tell, these are all open reference OTU. I know I need closed-reference OTU picking for Picrust.
I’ve attempted a few different methods (importing the rep_seq from step 1, rep_seq from step 4) using vsearch, etc.
However, there seems to be an issue with regards to which table to use. There appears to be some disconnect with regards to what the feature names are in the table and the ref_seqs.
I guess what I’m asking is… which seqs and which biom table should I be using for picrust and what steps do I need to take to get an OTU table formatted for picrust? I’m not sure which rep_seq or table would be analogous to the ones you get after DADA2/Deblur.
(kidding!)