recipes
for microbiome data analysisWork in progress!
This package is a wrapper around popular microbiome methods implemented packages such as phyloseq
or vegan
in the context of parsnip
, recipe
and workflows
in the tidymodels
ecosystem.
Specifically, tidymodels
allow for a unified tidy approach to data transformation, modelling and evaluation. It is also a way to keep provenance of analysis objects for reproducibility and sharing across researchers without having to slog through tons of code. A practitioner can print a recipe
model and examine data preprocessing steps, which is vital in microbiome data analysis. The desire to have a separate package to support phyloseq
objects is also because operations done on taxonomic data, such as tip_glom
or tax_glom
affect multiple layers of data, such as the phylogenetic tree, taxonomic table, etc. This is not readily available in vanilla recipes
This small package takes advantage of the readily available tools in parsnip
and recipe
to import custom methods in order to create a suite of step_
functions and parsnip
engines specific to microbiome research that can readily be used in the tidymodels
context.
The long term goal is to evolve around the tidymodels
ecosystem, with the really far-fetched goal of also support ExpressionSet
type objects in Bioconductor.
Install this work in progress using devtools
/remotes
package
remotes::install_github('quangnguyen1995/micrec')
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.