pipeline for meta analysis of different 16S regions

[mckin (Kingsley Apusiga)]

Hello Qiime2 Community!

I am new to microbiome analysis and I am trying to run a meta analysis.
I have 16S data (gastric and colorectal cancers) from different studies but mostly covering the V3 to V5 regions (all runs cover the V4 region) downloaded from NCBI. My goal is to identify microbes shared across these two cancers.

I am having a few challenges at the early stages of the analysis.

My questions are:
Should I trim to the V4 region for all runs? (I have already trimmed off all adaptors with Trimmomatic) and how do I perform such trimming? Do I leave the V4 data untouched whiles trimming a specific number of bases (depending on the region) off the start of V3V4 sequences?

My initial thinking is to process the different study data to be nearly identical (if possible) and then run dada2/deblur to produce individual tables which I will merge for further analysis. Is this appropriate?

What is the recommended approach to conduct such a study?

Thank you

[mckin (Kingsley Apusiga)]

(post deleted by author)

[colinvwood (Colin Wood)]

Hello @mckin,

I found some helpful posts related to these questions here and here. Let us know if these answer your question.

[mckin (Kingsley Apusiga)]

Hello @colinvwood,

Thank you for pointing me in the right direction with those posts addressing similar questions.
I have resolved to treat every specific study data independently but with similar parameters and then merge the tables afterwards. I will include various columns/variables as confounders in downstream analysis to help account for batch effects.
I will surely be here to ask further questions when they come up.