Rarefaction and normalisation

User Support
1

Dear Sir/ Madam,
I have a few questions about analysis pipeline in QIIME2.
I followed MOVING PICTURE tutorial.
I want to evaluate relative abundance of my data.
Apparently, rarefying is an important part in analyzing. I found two type of rarefying: 1. for feature-table 2. alpha rarefaction
To consider relative abundance of OTUs, do I need to carry out the both rarefaction methods mentioned above?
If not, do I need to rarefy my feature-table then try to provide a taxonomy barplot? If I do that I will lose statistical analysis which they are being calculated in alpha, beta and alpha rarefaction plot steps! looks confusing!

To summary, following the alpha, beta and alpha rarefaction steps are necessary to reach the relative abundance of OTUs?
To do this, do I have to rarefy feature-table (without analyzing alpha, beta and rarefaction) then plot my data taxonomy? If yes, another question raises. By rarefying table, all samples become at the same number, so how possible does it give me the actual counts of each sample while in previous step the counts have been at the same value?

2

alpha-rarefaction is more of a diagnostic. It tells you how the level of rarefaction impacts alpha diversity estimates. See the tutorials for some applied examples.

You use the results of alpha-rarefaction to diagnose what sampling depth you want to use prior to running core-metrics (which performs rarefaction, alpha diversity, and beta diversity all in one pipeline).

taxonomy barplots are different. You do not need to normalize prior to visualizing what species are present, but it is fine to use a normalized table too.

3

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