Hi all,
I'm excited to share a new community plugin from my team, q2-boots. q2-boots is designed to match the interface of q2-diversity, but rather than running a single step of rarefying a feature table, the actions integrate n iterations of resampling the feature table, with replacement (bootstrapping) or without replacement (rarefying). The requested diversity metrics are then computed on each resampled feature table, and the resulting alpha diversity vectors and/or beta diversity distance matrices are averaged and returned.
q2-diversity has had the alpha-rarefaction and beta-rarefaction Actions for a long time, but since the outputs are Visualizations, rather than Artifacts, integrating resampled diversity metric results in downstream analyses has required additional work on the part of the user. The Artifacts returned from q2-boots, on the other hand, are of the same semantic types as those returned from the corresponding Actions in q2-diversity, so can be used in any downstream analyses that support those types.
We were motivated to develop q2-boots by two recent papers, the latter of which suggests that rarefaction is "currently the best approach to control for uneven sequencing effort in amplicon sequence analyses". In addition to this being a more representative approach for computing diversity metrics (relative to the most widely used q2-diversity workflows), q2-boots now provides a straight-forward approach for generating data that more complex approaches can be benchmarked against. (It's worth noting that the second paper referenced above didn't compare bootstrapping versus rarefaction, but that would straight-forward to do with q2-boots - if anyone is looking for a methods comparison project, that would be a good one!)
q2-boots is in beta release, meaning that it is sufficiently validated and ready for real-world use, but interfaces (e.g., command names, which parameters are optional versus required, etc) are still subject to change as we get feedback from the user community.
Want to learn more? Check out the q2-boots pre-print on arXiv.
Enjoy!