Alternate to greengenes: Silva ?

[jaysiddharth]

Hi,

I was wondering if there was any way one can use picrust without using greengenes.

We have been using silva alignments and taxonomy as we have found it to be of much better quality.

With the option now available within Mothur to have picrust format (http://www.mothur.org/wiki/Mothur_v.1.33.0) I was hoping if the issue of using alternate reference databases can be addressed.

Kind regards

Jay S

[Morgan Langille]

Hi Jay, 

Did you happen to see this post? http://www.mothur.org/forum/viewtopic.php?f=10&t=2734&p=7642&hilit=picrust#p7642

Also, as you already linked to, Mothur will be able to generate BIOM files that contain GG OTU ids so that PICRUSt can use this as input. This will be available in Mothur v1.33, which I don't believe has been released yet, but the documentation illustrates how this would be done (see picrust sub heading on this page): http://www.mothur.org/wiki/Make.biom 

I believe the best solution is to wait until the version of Mothur is released. 

However, I should mention that PICRUSt was designed to be able to handle other types of marker gene trees (e.g. SILVA, or even a non-16S tree). This would require having a mapping from tips in the new tree that match sequenced genomes (we currently use IMG genome ids). Then precalculated internal PICRUSt files could be constructed for just your user table (fast) or all tips in the tree (slow, but usable by other OTU tables). This tutorial outlines the commands in PICRUSt to do this: http://picrust.github.io/picrust/tutorials/genome_prediction.html#genome-prediction-tutorial

Good luck, and if you are successful in getting PICRUSt predictions from Mothur OTU tables, please let us know so we can document it on the PICRUSt web page. 

Thanks,

Morgan Langille

Assistant Professor & Canada Research Chair Candidate

Department of Pharmacology

Dalhousie University
http://morganlangille.com

--
You received this message because you are subscribed to the Google Groups "picrust-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to picrust-user...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

[Indugu Nagaraju]

Dear Dr. Morgan

I got strcuk with PICRUSt in QIIME to predict function of 16s metagenome data based on Silva database reference sequence. 

I have also gone through the link you provided  http://picrust.github.io/picrust/tutorials/genome_prediction.html#genome-prediction-tutorial.

Based on pick_closed_reference_otus.py I hava generated otu table (attached).

Thank you

Sincerely

Nagaraju

[Joe James]

So can we use this OTU table if we did our 16S annotation in Silva? Our 16S amplicons were annotated much better using Silva than greengenes.

Thank,

Joe

[Jesse Zaneveld]

Hi Joe,

So right now there is not a built-in mechanism for doing metagenome predictions in Silva out of the box.  It's conceptually simple (one can use any trait/tree combination to generate the precalculated tables for PICRUSt if you have mappings between the trait table and the tree tips). In practice the step of mapping genomes to trees is often much more time-consuming that you would expect. So my recommendation for now would be to try out the predictions in greengenes, at least initially.  Feedback on which features are useful is definitely valuable though, and so I will definitely keep your case in mind. 

Best,
Jesse

p.s. for my own information, would you mind passing on the stats on your greengenes vs. silva mappings (and which versions)?  I'd be interested to see what you found. 

To post to this group, send email to picrus...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Maria Camila]

Hi all!

I see this issue is from 2014 so I was wondering if by now anybody has successfully run PICRUSt with Silva database as I want to attempt to do so. Also I wanted to ask if anybody has run PICRUSt with an ASV table (amplicon sequence variant) instead of an OTU table.

I just want to know how long does it take to set PICRUSt to use silva instead of greengenes or if there is now a simpler way of doing it with 16S data instead of doing the genome prediction step. 

Cheers.

Camila

[Gavin Douglas]

Hi Camila,

I’ve been working on a new version of PICRUSt here: https://github.com/picrust/picrust2

It’s currently in beta release, but it can be used with an ASV table (it essentially is an easier and faster way of running the genome prediction step with your own dataset).

Best,

Gavin

[Aqleem Abbas]

Dear Sir,

Thanks for your work on PICRUST. I have worked on qiime 2 and spliced my paired sequences of 16S with dada2 and got OTU table and Sequences table. Now I want to import these into PICRUST for further analysis. Please provide me step by step commands so that I import these into picrust and will do analysis. 

Thanks

Aqleem Abbas

Plant Pathology

HZAU China

[Gavin Douglas]

Hi,

You can use the PICRUSt2 plugin with QIIME2: https://github.com/picrust/picrust2/wiki/q2-picrust2-Tutorial

Gavin

To post to this group, send email to picrus...@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

[Aqleem Abbas]

Dear Sir,

Which file I need to upload in picrust from Qiime 2 side? I need step by step guidance

Thanks

[Gavin Douglas]

Please take a look at the tutorial I linked below - you can directly use the QIIME2 artifacts you already generated with DADA2 in QIIME2.

Gavin

[Aqleem Abbas]

Dear Sir,

Thank You. I have one more question. What is the difference between qiime2 and Picrust? Can we use picrust for fungal ITS analysis? Please clarify about the usage of both software?

Thank You

Aqleem Abbas.

[Gavin Douglas]

Hi there,

PICRUSt2 is a tool for functional prediciton from amplicon sequences. QIIME2 is a framework for using many different bioinformatics tools - including some features of PICRUSt2 currently. The database for ITS amplicons isn’t available yet, but it is on my to-do list. I think maybe you’re confused about what QIIME2 plugins are in general, which you can read about here: https://docs.qiime2.org/2018.11/concepts/

I shared the link to the q2-picrust2 plugin tutorial previously with you - if you have any specific questions I’ll be happy to respond.

Gavin

[Aqleem Abbas]

Dear Sir

Thank you for your message. Before exporting biome table to picrust there is need to close reference OTU picking. Why we need to do this step?

Thanks

[Gavin Douglas]

Closed-ref OTU picking was needed for PICRUSt1, but this isn’t required for PICRUSt2, which you can read about here:https://github.com/picrust/picrust2/wiki. Make sure you’re looking at this page if you want to use the PICRUSt2 QIIME2 plugin: https://github.com/picrust/picrust2/wiki/q2-picrust2-Tutorial

Gavin

[Aqleem Abbas]

Dear Sir,

Thank You. 

[Aqleem Abbas]

Dear Sir,

Do you have some idea about the trimming of paired reads 16 S clean reads(No primers)( 2* 300)  using DADA2 in qiime? I did not trim the forward and reverse reads as a result the highest sequence count in my sample was 3400 and lowest was 796, However when I trim and no of sequence count in my table.qzv has increased. The highest goest upto 20000 and lowest goes upto 10000. Now which is best threshold? What is going on here? I have talked to qiime2 forum but they did not reply in a good way. I could not understand. My demux reads were highest 80000 and lowest were 40000. 

Thanks