I got my raw data generated from QIIME1 pipeline, which the barcode and primer have been trimmed, and low quality sequence has been discarded (I think it is not suitable for DADA2).
So after I inputed it to QIIME2, I performed the vsearch cluster-features-de-novo for OTU table and representative sequences. I further performed the feature-classifier classify-sklearn by SILVA database.
My question is, I have removed the unwanted Taxa from my OTU table (and some quality control like chimera removal), if I want to performed the closed reference clustering for PICRUSt, should I used my “clean” OTU table or the one just demultiplexed?
I don’t know if it’s ok to performed the closed reference clustering based on the table that has been de novo clustered
Any suggestion will be help!!
Thank you very much!!
Hi @Lennon_Lee, I would recommend starting with the closed-reference OTU picking step, rather than closed-reference following de novo clustering. Otherwise you may end up clustering some sequences together in the de novo step which may have hit different reference sequences (and therefore been assigned to different OTUs).
That said, have you seen the new PICRUSt2 plugin? This will allow you to use PICRUSt2 without a closed-reference clustering step which improves your resolution, and importantly doesn’t discard sequences that aren’t high percent identity matches to the reference database. If possible, I would use this - there is a tutorial linked from the top-right of that page.
Also, one other bit of unsolicited advice: I recommend processing all of your data in QIIME 2, rather than starting with QIIME 1 and then moving to QIIME 2. This will simplify the description of your bioinformatics workflow, which can be important for your own reproducibility (e.g., if you need to redo parts of this analysis some day), and if/when you write a paper on your study.
Hope this helps!
I didn’t noticed that we have PICRUSt2 plugin!!
Thanks for all your advices!!
An off-topic reply has been split into a new topic: How to install a plugin in qiime 2 docker container
Please keep replies on-topic in the future.
This topic was automatically closed 31 days after the last reply. New replies are no longer allowed.