Merging Feature Tables - Should I also Merge Taxonomy or does it need to be re-run?

User Support
1

Hi,

I have two feature tables for samples that were sequenced together, but because of computational issues for DADA2 I split them into two logical groups based on a metadata category. That was all successful and I've been able to do some downstream analyses on the two separate groups. However, I'm finding that I need to do some analyses for the entire set together.

I've successfully merged the two feature tables and the rep-seqs files. I'm wondering, what would be the best practice now...can I use merge-taxa-data to merge the two taxonomies together into one cohesive taxonomy for the entire data set? Or do I need to re-run that taxonomic classify step (which will take over a week for this data set and causes some serious grumping from our supercomputer squad)?

If I use merge-taxa-data will it know how to merge the feature IDs to match the merged feature tables and rep-seqs?

Thanks!
Alicia Reigel

2

sure. If the sequences are identical, reference is identical, and classification method/parameters are identical, you should get the same result for the same sequence, so merge away.

No need for grumping of any kind! just qiime feature-table merge-taxa :lemon:

Yep! You can read the help docs for that method to learn about what happens when the two taxa disagree on the taxonomy for a feature ID.

Good luck!

3

AWESOME! Thank you! This is super helpful and should be way more efficient than rerunning the taxonomy step since they were all treated the same throughout the process!