qiime2 studio vs q2cli result

#1

Hi everyone,
I went through qiime2 documents to learn about basic qiime2 commands. Since, I never used commandline before, i struggle to import files. In addition, I am getting ION TORRENT files from my collaborator which are difficult to work with qiime2. Hence, I started using qiime2 studio. Studio is simple and intuitive. I hope qiime2 developer team will make it more efficient in the future.
Having said that I was analysing moving picture tutorial data to compare result between studio and q2cli version. I found that studio results are different for dada2 analysis and beyond than q2cli. Why is it so? Am I doing something wrong? I am uploading table from dada2 analysis in studio (1st file) and q2cli (2nd file) for suggestion.
table.qzv (350.7 KB)
table.qzv (350.7 KB)

Thank you
Regards
Sudhir

#2

These table summaries appear identical to me - can you please clarify what differences you are seeing?

#3

Hi Mathew,
Thank you for your quick reply.
I see the number of feature counts different for studio and q2cli.
It seems that studio is giving more counts per feature as compare to q2cli table.
Because of this sampling deepth is higher in studio compare to q2cli.
Please compare this studio file table.qzv (350.7 KB) with q2cli table https://view.qiime2.org/visualization/?type=html&src=https%3A%2F%2Fdocs.qiime2.org%2F2019.7%2Fdata%2Ftutorials%2Fmoving-pictures%2Ftable.qzv
Because of this my diversity plots are looking different.
Thank you

#4

Hi @Sky23 — thanks for sharing new QZVs (the first ones you sent were identical).

The explanation is simple, and has nothing to do with q2cli vs q2studio. The explanation is this:

You used two different versions of q2-dada2 (2019.1 vs 2019.7), which, uses two different versions of DADA2. These produced slightly different ASV results, but should more-or-less be the same.

Hope that helps!

#5

thank you so much mathew. Now it make sense.
I wish qiime2 studio has the same capabilities as q2cli in the future.
Studio is really a game changer for non-bioinformatics person like me.

2 Likes