gneiss icon indicating copy to clipboard operation
gneiss copied to clipboard

Median and dispersion estimators

Open mortonjt opened this issue 8 years ago • 0 comments

This is building off of the differential abundance estimators used in ALDEx2

This may help with accounting for pseudocounts.

Basically, this would bootstrap a Dirichlet prior, and take the median and dispersion as a proxy for the mean and variance for the estimator.

Ultimately, this will allow us to account for variances in the low counts.

CC @ggloor

mortonjt avatar Jun 27 '17 23:06 mortonjt