my rarefaction curve looks like a ladder, and I have a sample that doesn't reach the bottom of the graph. what should I do, conceptual help please.
What happens with the treatment that does not reach the max-deph if I cut it at 114000 approx.?
And if I cut at approx 9500, would that be correct?
Hi @Leytoncito,
Thanks for reaching out! :qiime2:
Happy to discuss your questions and provide some resources that may be helpful in your conceptual understanding.
This 'ladder' structure that you are referring to is actually what we expect to see in our alpha and beta rarefaction plots - there will be areas of stability as well as instability (the relatively flat segments, and the segments where we see a change in the Y-value i.e. the alpha diversity measure, respectively). The extent to which we see these structures is highly dependent on the data itself, as well as the chosen sampling depth (what you used when actually generating the data subsequently used to create this visualization).
The sample that doesn't reach the bottom of your alpha rarefaction plot (Valle de la Luna) indicates that between a sequencing depth of 90k to roughly 107k, the observed_features metric was around 1200.
You won't see any data past your chosen max-depth value. I'll discuss best practices for selecting a reasonable max-depth for your data below.
This depends on what you are wanting to look at in your data. For alpha rarefaction, you want to select a max-depth that is rougly 2-3x bigger than your chosen sampling depth.
When selecting a sampling depth, it's best to examine your feature table summary and use that to determine a good depth for your data (if this isn't something you've already done). Typically, you want to aim for a sampling depth that is as high as possible (thus retaining more sequences per sample) while excluding as few samples as possible. That being said, this is highly dependent on what your study entails and what you are specifically trying to test.
When examining these alpha rarefaction plots, you can also gauge if you've chosen a sampling depth that accurately reflects your data if you are seeing stability in the alpha diversity metric at that sequencing depth (in your case, that would be observed_features).
This is a really helpful tutorial video that discusses sampling depth in detail - this may also help further clarify the process of selecting an even sampling depth and analyzing your subsequent alpha rarefaction visualizations.
Hope this helps!
Cheers 
This topic was automatically closed 31 days after the last reply. New replies are no longer allowed.
