I import all samples and generate the table and tree for all samples in qiime2. However, for the core-metrics-phylogenetic analysis, I only would like to retain certain samples. after Identifier-based filtering to filter the feature table, I use the filtered table and the original tree file for the core-metric. Were I supposed to do this? Are the distance matrix recalculated? or those number are still from the original calculation with only samples removed?
Hello Rachel,
Yes, I think this is all good!
When you filtered the table to only include specific samples, the features/OTUs in that table are probably the same. As such, the tree is still valid and any phylogenetic metrics you calculate are still valid for that tree.
(Distance metrics that use phylogenetic info like UniFrac will use the full tree, but but because the tree is the same, the sample-to-sample distances are the same even when you use only some of the samples.)
Colin
P.S. If certain samples you removed have features/OTUs you don’t want in your tree, then you could build a new tree that omits these features. 
This topic was automatically closed 31 days after the last reply. New replies are no longer allowed.