Ion 16S Metagenomics Kit for Ion Torrent?

[JenB]

Hello,

Has anyone in this forum tried out the 16S Metagenomics Kit for Ion Torrent and then if so, did you try to analyze the data with Qiime?

I previously had analyzed Ion Torrent data using Qiime and the Brazilian Microbiome Project pipeline: 

however this kit makes use of 7 primers and sequencing 7 variable regions.

• 16S primer set V2-4-8 (300 µL)
• 16S primer set V3-6,7-9 (300 µL)

 So my previous pipeline won't quite work with this.  Has anyone tried this and used Qiime for the analysis?  

Life Tech delivers a software pipeline along with this for the analysis but I am curious to find someone that has tried it out and was able to get good results?

thanks,

Jen

[Daniel McDonald]

Hi Jen,

I haven't touched Ion Torrent data. A quick poke on the forum though shows this posting:

https://groups.google.com/forum/#!topic/qiime-forum/NEotQnTQlQo

Does that help at all? If not, we can dig deeper on it.

Best,

Daniel

--

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

[JenB]

Hi Daniel,

Thank you for responding.  I actually read a lot of the Ion Torrent posts earlier this year when I was working out my pipeline for some preliminary Torrent data that we had however now we have this new 16S Metagenomics Kit data from Ion Torrent that provides reads from 7 of the V regions and I was wondernig if anyone had used the kit.  

It is pretty new so it is probably unlikely but I thought that I would ask the question.

Info about the kit is here:

https://www.lifetechnologies.com/order/catalog/product/A26216

it sequences 7 variable regions of the 16S gene.

Jen

[Daniel McDonald]

Hi Jen,

I don't have any specific experience and I'm not aware of anyone else who has. I'd be particularly concerned about the potential for bias from reads from the different regions. For instance, Lozupone et al 2013 showed a striking difference between HMP v1-3 and HMP v3-5. 

My guess is that the best approach would be to demultiplex each primer separately, and do closed reference OTU picking to minimize bias from the different variable regions. Does that basic strategy seem to make sense?

Best,

Daniel

[JenB]

Hi Daniel,
Thank you for the update.  Yes, this makes sense however I just found out that Life Tech does not publish the primers to the kit thus I would not know which reads come from which primer.  I am looking into how I could try to find the primers myself in the reads but out of curiosity, do you have any suggestions on how I could do this?

i have the fastq files for each of my samples with reads coming from the 7 different regions.  I just don't the primer sequences in them.

thoughts?

Thanks a bunch,

Jen

[Daniel McDonald]

You could align a subset of them with PyNAST against the Greengenes core set and see what regions are covered. I don't know if you need to do this though. What about demultiplexing each file separately, pick OTUs (closed reference) and then merging the resulting OTU tables?

Best,

Daniel

[JenB]

Hi Daniel,

Again thanks for the posts.  I am looking for any advice at all on what i could do with this data without knowing the primers.  Again, this is a set of reads that are derived from sequencing 7 different regions in the 16s gene.

The files are already separated, demultiplexed by sample so that issue is taken care of.  You mentioned pick OTUs (close reference) and then merging OTU tables.  Can you be more specific on what I could do?

I have fastq files for each sample (16 samples (1 sample per file) in total).  I dont know which region a read is from in each file because I don't know primer info.  The library is derived from pooled primers.  Is it possible to move forward with an analysis without having that information?

thanks,
Jen

[Daniel McDonald]

Hi Jen,

I believe you should be able to do the following:

$ split_libraries_fastq.py -i sample_1.fastq,sample_2.fastq,...,sample_N -m your_mapping_file.txt --sample_ids sample_1,sample_2,...sample_N -o slib

$ pick_closed_reference_otus.py -i slib/seqs.fna -o otus -r path/to/greengenes_otus/rep_set/97_otus.fna -t path/to/greengenes_otus/taxonomy/97_otus.txt

The first command should process your per-sample fastqs. It is important to make sure that the order of the fastqs specified corresponds exactly with the order of the sample ID specified, and that each sample ID exists in your mapping file.

The second command should do the closed reference OTU picking. Please note that you'll need to fill in the path to the Greengenes OTUs.

I hope this helps, please let us know!

Best,

Daniel

To unsubscribe from this group and stop receiving emails from it, send an email to qiime-forum+unsubscribe@googlegroups.com.

[Daniel McDonald]

Hi Jen,

I apologize, but I may have led you astray. Instead of the split_libraries_fastq.py command above, you instead want to first use add_qiime_labels.py. You can then quality filter (if you'd like) using split_libraries_fastq.py. This is outlined more in this post. 

Best,

Daniel