v2.0.0 better, faster, stable-r

After so much helpful feedback from users, the radEmu team is delighted to share our next major release of radEmu.

This is the same rigorous and robust differential abundances method as always, and is now faster, more stable, and better documented. Sarah has created some incredible vignettes, including

• parallelization
• setting a reference taxon (eg for those with spike-ins)
• cluster correlated data (is useful for many experimental designs including, repeated measurements/longitudinal data, cage/batch effects)
• data in both phyloseq and TreeSummarizedExperiment formats
  and David has been in the weeds of speed and memory profiling, tweaking defaults and streamlining the code.

One thing you may notice is that we now require test_kj by default. This will be prevent emuFit() from running every score test without your explicit request, saving some time for users who, errrr, skipped the documentation. We are happy to help with backwards compatibility issues, but believe that this is likely to be a good thing for almost all uses.

Something essential to making this new release possible was constructive community feedback. If there's something you'd like to see, please let us know. We are very committed to making estimating differential abundance a joyous statistical experience for you, and continue to do our best to support our users. Many thanks to those of you who reached out with questions and suggestions.

We hope you enjoy v2.0.0 of radEmu!

Wishing you fast converging estimates and highly-powered tests,

Your radEmu team ❤️

v1.0.0 to accompany arXiv

This is our first official release of radEmu!

radEmu is a method for estimating and testing differential abundance of microbial taxa across covariate groups, without the need for pseudocounts, log-ratio transformations, priors, a single binary covariate, etc. It handles count and non-count data, making it relevant to both modeling shotgun sequencing depths as well as amplicon/ASV counts. Critically, it estimates a meaningful biological effect even in the presence of differential detection and uneven sequencing.

Many thanks to the attendees of STAMPS 2023 for their fantastic feedback that helped us improve the software and input/output.

Software authors: David S Clausen, Sarah Teichman and Amy D Willis
Methods authors: David S Clausen and Amy D Willis