qiime feature-table merge and the other merge sub-commands only allow for merging two tables. Given that dada2 is supposed to be run separately on each sequencing run, a user working on a dataset consisting of many sequencing runs must iteratively merge all feature tables after running dada2 on each sequencing run. If there are many sequencing runs (eg., dozens for the dataset that I’m working on), the iterative merging leads to a lot of redundant qiime feature-table merge commands, or a somewhat complicated bash script to handle a dynamic number of feature tables to iteratively merge.
It would be great if the merge sub-commands could merge >2 tables. This shouldn’t be much extra code, and it will prevent the need for users to invent their own code for merging many tables. This is likely going to be a common issue for many people, given that many microbiome datasets are comprised of many sequencing runs.
Most definitely! This has been a pain-point we've been pretty unhappy with and we are currently working on what's needed to avoid it.
If only that was true
Here's some PRs I'm working on to enable variadic input if you want to keep up to date with the latest progress:
That doesn't get into actually allow feature-table merge to accept multiple inputs, but once these PRs are merged we'll be able to get started on that!
