what's the meaning of the result of "Alpha rarefaction "?

Dear all,
I alreay know how to do "Alpha rarefaction plotting", Here are my input and result.

qiime diversity alpha-rarefaction
--i-table /share/disk0/user/maj/YuZhang/micro_plastic_2020/16s/Download/table-dada2-3.qza
--i-phylogeny rooted-tree.qza
--p-max-depth 85554
--m-metadata-file metadata.tsv
--o-visualization alpha-rarefaction.qzv
--output-dir core-metrics-results


But,I dont know how to use the result. I mean, I know my sample will decrease at Sequencing Depth with 30000,but Idont know . But I don't know which parameter will be affected by this result . Alpha diversity?

My input about alpha diversity:

core-metrics-phylogenetic
--i-phylogeny rooted-tree.qza
--i-table /share/disk0/user/16s/Download/table-dada2-3.qza
--p-sampling-depth 30908
--m-metadata-file metadata.tsv
--output-dir core-metrics-results

Here are my feature table:

Thanks a lot,
Yu

Hi!
Based on these two plots, one can decide, which sequencing depth to use for rarefaction in core-metrics results.
For example, if you choose 10 000, you retain all your sample, but loose a lot of reads and potentially useful information about features/taxa, especially that occurs relatively rare, since lines still increase after 10 000. If you go for 30 000, you are still keeping your samples, but taking into account much more reads/rare taxa. If you go for 60 000, you will loose all samples that are below chosen depth. So you need to decide, what is better for the analysis - to keep all of the samples, or to sacrifice some samples to increase the number of sequences, that will be included in the analysis.
In my opinion, something around 30 000 will be fine.

1 Like

Thanks,you mean,the result of alpha rarefaction,determin which sample depth should I chose in core-metrics-phylogenetic?
Meanwhile, I have another question.What the deference between alpha deversity and alpha-phylogenetic ,and core-metrics and core-metrics-phylogenetic? I don’t know which should I chose to cabulate the diversity. eg. shannon,simption,faith_pd …Does alpha-phylogenetic only return the result of faith_pd ? And alpha deversity only return the result of shannon,simption…?

That’s right

Alpha diversity - by this plugin, you can compute any alpha diversity metric by your choice, excluding alpha diversity metrics, that require a phylogenetic tree (faith_pd)
Alpha diversity phylogenetic - you can compute any alpha diversity metric by your choice, that require a phylogenetic tree (faith_pd, are there any more?).

The same is right for core-metrics and core metrics phylogenetic. In first case, you don’t need a phylogenetic tree and receive as an output alpha and beta diversity metrics that do not require a phylogenetic tree, in second you will receive both - metrics, that requite and do not require a phylogenetic tree.

So, unless you absolutely sure, that you do not need metrics, requiring a phylogenetic tree, it is better to perform core metrics phylogenetic to obtain both, phylogenetic and non-core metrics phylogenetic alpha and beta diversity metrics.

2 Likes

Thanks for your reply. In sort, core metrics phylogenetic includes core metrics and phylogenetic such asfaith_pd_vector.qza, so I can use core metrics phylogenetic only. If I want other diversity, I can use Alpha diversity to calculate. Is it right? And shannon, simpon, observed-otu and chao1 not rely on Alpha diversity phylogenetic ,right?

Yes, that’s right.
I would construct a tree, perform a core-metrics-phylogenetic, and use alpha-diversity for any additional alpha d. metric if necessary

1 Like

Thanks,I think the same as you.

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