Comparing deblurred sequences from different datasets

Hello,

I’d like to run a meta-analysis or comparison using my data and a dataset deposited in Qiita but using the deblurred sequences from both sets. I’m wondering at which step or what files should I merge. The example meta-analysis in Qiita works on otu tables from closed-reference-otu picking step only. Is it possible to perform meta-analysis for deblurred datasets in Qiita? Or should I just download the trimmed-demultiplexed fasta, combine, and perform otu-picking again, and assign taxonomy in Qiime?

Thank you!

1 Like

Hi @maria,

It is possible to perform non-phylogenetic meta-analyses using Deblur in Qiita right now. If you’re continuing to have issues, would it be able to send an email to "[email protected]"?

If you’d like to proceed outside of Qiita, you should merge against the reference-hit.biom table. I recommend using a common trim length, and additionally, to set --p-min-reads=1 when you run Deblur on your data to ensure each sample is processed truly independently (this parameter filters out low abundance reads within a single execution of Deblur across samples). After merge, you may wish to filter out low abundance reads like singletons.

Best,
Daniel

1 Like

Hi, @wasade,

I have a similar problem. I looked for the code you posted on the github https://github.com/cuttlefishh/deblur
on how to use deblur to process the reads.
I have carefully read the data processing on the paper published in nature. https://www.nature.com/articles/nature24621
It mentioned trimming all reads to 90bp or 100bp.
Here is the situation. I also want to trim the reads downloaded from EMP project. May I know how to trim them to the same length? Just --p-trim-length 100 in qiime deblur denoise-16S procedure? I am confused.
Thanks in advance.

Hi @Lu_Yang,

That’s correct, you can trim the reads by specifying --p-trim-length, and an analogous procedure was performed for the EMP.

Best,
Daniel

1 Like

Hi, @wasade,
I see. Thanks so much.
Best.

1 Like

This topic was automatically closed 31 days after the last reply. New replies are no longer allowed.