QIIME in Illumina BaseSpace app

[Melissa Manus]

Hello, has anyone used the QIIME apps via Illumina BaseSpace? I am looking into using them for a quick analysis of MiSeq samples (ultimately want to assess alpha diversity via Fath's PD and/or OTUs, and beta diversity via UniFrac) but have never seen these apps cited in any publications. Are the apps comparable to actually working through a QIIME pipeline? I am curious how the quality of the analysis differs, and if this is seen as a legitimate way to analyze data. Thanks so much!

[Yoshiki Vázquez Baeza]

Hello Melissa, 

Yes, these are legitimate! We authored and maintain these, so let us know if you find any problems with them. Note that the apps themselves are intended to give you a quick way to generate OTU tables, taxa summaries, alpha diversity and beta diversity data/plots. However they are not intended to give you the fine parameter tuning that's possible with the command line scripts. Specially BaseSpace is useful to bypass the need for a compute cluster, since all this is resolved through the platform itself.

Thanks!

Yoshiki.

[Melissa Manus]

Thanks so much, that's what I was hoping to hear!! I do have another question for you or anyone else with experience-- I'm working with the QIIME visualization app right now, and my analysis keeps aborting. I know that the error is related to the rarefaction depth that I am choosing (see below). However, this is just a test run of the app, meaning there is only one sample in the preprocessing input file, and that same one sample is in my mapping file. Could you briefly explain the rarefaction depth-- is that a percentage, or an actual integer chosen from the table-summary.txt file that the preprocessing app creates? I chose a rarefaction depth smaller than the "counts/sample detail" for my one sample in the .txt file (that number is 9820, so I chose 7000 for rarefaction to see what would happen). Thank you!

Error in make_emperor.py: Due to the variation explained, Emperor could not plot at least 3 axes, check the input files to ensure that the percent explained is greater than 0.01 in at least three axes.

[Yoshiki Vázquez Baeza]

Hello Melissa,

The problem you are seeing is not directly related to the rarefaction depth, it's actually caused because you have only one sample in your study. Emperor, our 3D PCoA plot viewer, has a hard limitation on not being able to display ordinations that can be explained with less than 3 axes of variation. I wouldn't worry too much about this, if you still want to give this a shot, I would suggest that you use ~3-5 samples.

Rarefaction as implemented in single_rarefaction.py and core_diversity_analyses.py, requires an integer i.e. the rarefaction depth, this is the value that will be used to randomly subsample all your samples so they have exactly that amount of counts. One final note is that samples with less counts that your  specified rarefaction depth will be thrown out.

Thanks!

Yoshiki.

[Melissa Manus]

I understand now, thank you!

I'm now getting an error that the sample names in the mapping file do not match the sample names in the QIIME preprocessing input files. I see the problem is that the names in the preprocessing file include a " - ", but that specific character is not allowed in the mapping file (it gets flagged by Keemei). The names in the mapping file include a " ." instead, but I can't change them using the disallowed " - " character. Is there a way to change the sample names on BaseSpace so that they match the mapping file? Thanks!

[Yoshiki Vázquez Baeza]

Hello Melissa,

I think I understand what's going on, this can be traced back to the fact that characters considered valid by QIIME and characters considered valid by BaseSpace are slightly different. Thus only allowing us to use alphanumeric characters (characters that are valid both in QIIME and BaseSpace). I think the solution to this problem would be to change the name of your samples on the Illumina side of things, assuming of course this is not too problematic. I've never really dealt with that side of the sample processing, so I can't help that much here.

Thanks!

Yoshiki.

[Melissa Manus]

Yes, that's the problem. Well thank you, I might repost this question to the forum to see if anyone has attempted to change things on the Illumina side. Thanks!

[Sonia García-Carpintero]

Hi Melissa, my name is Sonia. I written about qiime basespace app because I see you in the same situation than me. I make qiime apps basespace for processing 245 samples in the same time and 3 days later continue running. Are you a similar situation sometime? How much samples Can I do in the same time?. My study is a big study with human population, I sequencing 16S with Illumina Miseq and now I need processed this samples with qiime. But I dont know How use the originals qiime.  Thanks you for your help

[Lisa Weissenburger-Moser]

Hi Yoshiki,
I was wondering if you could tell me the defaults that are used with qiime app in basespace? I'd like to know what is going on in the pipeline since you are not specifying anything like you do in regular qiime.
Thank you!
Lisa

[Yoshiki Vázquez Baeza]

Hello Lisa,

All the scripts used in the BaseSpace app are using the defaults as specified by each of the individual scripts:

pick_closed_reference_otus.py (default is to use 97% identity with the Greengenes database 13_8 release).

multiple_split_libraries_fastq.py (underneath the hood it uses split_libraries_fastq.py and its defaults).

core_diversity_analyses.py (rarefies at the user specified level and uses the Greengenes tree to compute the alpha and beta phylogenetic metrics).

If you are ever curious about the internals, all the source can be found here:

https://github.com/biocore/basespace-qiime

Hope this helps!

Thanks!

Yoshiki.

[Lisa Weissenburger-Moser]

Hi Yoshiki,
As you know I used Illumina's basespace apps, Qiime preprocessing and Qiime visualization.  I was wondering if you knew the minimum quality threshold value as well as the minimum read length that was used for the default?
Thanks!
Lisa

[Yoshiki Vázquez Baeza]

Hello Lisa,

The quality control parameters used are the defaults as provided by split_libraries_fastq.py. If you go to that documentation page, all the parameters are described, as well as their defaults. There's no one parameter that specifies the minimum read length, the consideration in this script is with regards to number of consecutive bases with high quality.

Thanks!

Yoshiki.

[Lisa Weissenburger-Moser]

Thank you!  Looks like it's 3 which isn't that great.

[Garrett Wilson]

Hi all, 

I started a QIIME Preprocessing run on Sept. 15 with 40 samples. Any idea how long it should take to process?

Thank you, 

[Jose]

Hi Garrett,

the time to process samples can vary depending on a number of things: the amount of diversity you have, sequencing depth, etc. So it's difficult to say exactly how long it would take for you, sorry we can't be of more help.

Jose

[chuanwu wang]

Hi Lisa,

What did you do to improve the quality of your seq ?

Thanks

Chuanwu

[Lisa Weissenburger-Moser]

 I used cutadapt.

[chuanwu wang]

Could you email me the full command you used for this?

Thank you

[Lisa Weissenburger-Moser]

It's a completely different program.  It's not through basespace.