Hi,
I have one last clarification, but given that it is on a different topic I decided to create its own topic.
For alpha diversity metrics, the calculations were calculated on rarefied, total abundance values correct, not relative abundance?
FeatureTable[Frequency] artifacts are used as input, so yes, it is frequency not relative frequency.
Whether or not the table was rarefied depends on the steps you performed — you will need to check the provenance on your results files with view.qiime2.org
I hope that helps!
1 Like
Yes it does, thank you.
Can I ask for your advice on relative abundance? I have seen on a few sites that alpha and beta diversity should not be calculated on relative abundance data, but on some papers they have used relative abundance to calculate both metrics and to perform the NMDS. So as per usual, now I am confused on whether I should perform relative frequency prior to calculating all metrics.
Additionally, I went ahead and calculated relative abundance using q_iime feature-table relative-frequency_ and tried to calculate alpha diversity but I get an error stating
"qiime diversity alpha
–i-table EcM-table-NS-rare-RA.qza
–p-metric simpson
–o-alpha-diversity simpson_RA_vector.qza
Plugin error from diversity:
Argument to parameter ‘table’ is not a subtype of FeatureTable[Frequency].
Debug info has been saved to /tmp/qiime2-q2cli-err-_5mga_lk.log
Calculating alpha diversity on relative abundance should not make a difference — depending on the metric (e.g., richness will be the same) — but it sounds unusual.
I have heard of some people using relative abundance as a normalization method for beta diversity analyses but I'm not sure what the consensus is on that.
For both of these, read this paper for more information.
QIIME 2 cannot use relative frequency tables for calculating diversity metrics! You need to use a FeatureTable[Frequency] artifact.
I hope that helps!
1 Like
This topic was automatically closed 31 days after the last reply. New replies are no longer allowed.