Help interpreting Chao1 rarefaction curves

Hello, I am very much a novice at QIIME (and bioinformatics more generally) and was wondering if anyone could please walk me through how to interpret QIIME 2's Chao1 rarefaction plots.

Here's a plot I generated based on sequences obtained from different O2 concentrations:

chao1.qzv (421.5 KB)

I essentially just want to describe how richness changes between my three O2 regimes.

Here are my basic questions:

  1. Do I interpret these as standard box and whiskers plots? Specifically, does each box denote interquartile ranges, with the horizontal line showing the median value, and the whiskers showing the maximum and minimum values?

  2. If #1 is correct, then is there a way for me to obtain the actual median values? Like, as calculated by QIIME, and not just me drawing a line to the y-axis?

  3. Using QIIME, is there a way to perform a Kruskal-Wallis test with a Benjamini-Hochberg correction on these Chao1 estimates the same way you can with other diversity metics (evenness, phylogenetic diversity, etc.)?

Overall, I know richness is lowest in the orange bin (labeled 3_below 1), but I would like to say quantitatively (or statistically) how much lower it is compared to the other bins. Does that make sense?

Any help is appreciated! Thank you.

Hi @DBM,
I think that this alpha rarefaction plot is not really what you are looking for in terms of comparing different O2 concentrations.

This plot is designed to show what a good sampling depth is, and not really comparing richness in different regimes Here is a video about rarefaction that might help. https://www.youtube.com/watch?v=g5BdGP4V5YA

I think what you are looking for is really alpha-group-significance: alpha-group-significance: Alpha diversity comparisons — QIIME 2 2023.5.1 documentation
This will give you big boxplots at only one sampling depth, and will contain statistics for comparing your groups.

Yes, it does, but it is highlighting alpha diversity at different sampling depths. It is not really comparing between groups.

This isn't not really available. The boxplots are bigger in alpha-group-significance so it is easier to roughly see the mean. Also from alpha group significance you can download the raw tsv of the data and calculate the median value for each group through that.

Yes! So what you are going to is generate the chao1 diversity vector through qiime diversity alpha and then run alpha-group-significance.

Overall, I think that this visualization isn't exactly what you need to answer your questions. I think the directions above should fix this!

Hope that helps!
:turtle:

1 Like

Hi Chloe,

Thank you so much, that was super helpful!

I have remaining question, though: I've run the core-metrics-phylogenetic method already, but it didn't seem to produce a chao1.qza file. Is there a way to do this? So far, I've only produced a chao1.qzv file (the one I attached above) using the qiime diversity alpha-rarefaction visualizer, but I can't input this file for the alpha-group-significance method.

Thanks again!

Hi @DBM,
You will need to run qiime diversity alpha to get a chao1 diversity vector then you should be able to hand the output of that to alpha-group-significance.

Hope that helps!
:turtle:

1 Like

Hi Chloe,

Yes, that worked, thank you so much!

2 Likes

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