How to validate my results

Dear All,

I was wondering if someone could walk me through how I can validate my results. I have performed a study using qiime 2 pipeline. I have gone through every step of movie pictures tutorial and now I am willing to see if my results including beta and alpha analysis, taxonomic analysis are valid.

I was told the only way is to compare my results to the results of another study.

Any help would be greatly appreciated.

Hi @ptalebic,

This is a complex question :), or at least more complex than it should be.

The simple answer for validation is to look for a study with a similar design, annotate it with a similar pipeline, and see if you get similar results. Qiita is a database of studies. Alternatively, if have a paper in mind and the data isn't avaliable through qiita, there's often a description of data avaliability. There have been some discussions of getting the data from EBI/SRA into :qiime2:,so maybe look there.

However, there's a set of secondary issues. The observed community is sensitive to extraction method, storage, hypervariable region, sequencing method, etc, etc, etc. Unfortunately, microbiome research hasn't hit a point with standardized protocols, and so anything you do is going to be affected by the wetlab and it may not be a direct comparison.

A second concern is the intrasepcies variation in the microbiome. International comparisons find big differences in the microbiome (American Gut Project, Teddy project). So, depending on what you pick as your reference study, you may not see the same response.

This leaves out issues around sample size, potential confounding, and what "replication" means. And so, depending on what you're looking at, you may or may not be able to replicate.

Best,
Justine

4 Likes

Thanks Justine for your great explanation.

1 Like