How can I best analyze the 16s amplicon sequencing data obtained from multiple runs? Do I have to run all the sequencing files in Qiime again? Are there any tutorials available?
Hello Jyoti,
Yes you can merge different sequencing outputs and analyze them together with QIIME2, but you have to do it before transforming the data into relative abundances. In QIIME2 you can do it after demultiplexing and denoising steps, later the frequency tables and representative sequences of your two datasets can be merged and from that point you can continue as if they were one.
On QIIME2 webstite there is this tutorial addressing a similar situation.
https://docs.qiime2.org/2021.8/tutorials/fmt/#merging-denoised-data
But if you already produced your feature tables from the different runs (hopefully using the same parameters for all of them) in R using Phyloseq package you can merge the different outputs with the merge_phyloseq command, and from that point performe the rest of your analyses.
Best wishes,
David
The best and most correct way to proceed would be reprocessing all the raw-data from different sequencing run with same parameters and together. The OTUs will surely change. With merging two separate runs, you would be forcing the two datasets to appear similar. Reanalysis would take only a few hours anyway.
Yes Jyoti Singh , you can do it. I did it for compare the microbiota composition from different studies:Article Cross-Regional View of Functional and Taxonomic Microbiota C...
.To diminish and reduce the data noise I re- analize all the data coming from differente studies using the same pipeline and parameters, later I scaled proportionally the relatives abundances to equalize the data distribuition, and then I normalized the scaled data using quantile normalization. Alternativelly, you can look information about to "reduce batch effects between data". Good luck!
Thank you Daniel A. Medina Abhijeet Singh David Rosado-Porto for your suggestions.
Beware of batch effect, and deal with properly during analysis.
Article Batch effects account for the main findings of an in utero h...
- Badges
- Science method