bayestax2df: Converts the output of DADA2's assignTaxonomy, which implements a naive bayesian classifier, into a dataframe compatible with the algorithms used in ensembleTax

Description

Converts the output of DADA2's assignTaxonomy, which implements a naive bayesian classifier, into a dataframe compatible with the algorithms used in ensembleTax

Usage

bayestax2df(
  tt,
  db = "pr2",
  ranks = NULL,
  boot = 0,
  rubric = NULL,
  return.conf = FALSE
)

Arguments

The taxonomy table output by DADA2's assignTaxonomy function.

The database you ran assignTaxonomy against. Either "pr2", "silva", "rdp", or "gg" are supported. You may set to NULL and include a character vector of rank (column) names for other databases.

ranks

NULL, or a character vector of column names if db is set to NULL

boot

The bootstrap threshold below which taxonomic assignments should be set to NA. This can also be done with DADA2's assignTaxonomy but is included here for convenience.

rubric

NULL, or a DNAStringSet (see Biostrings package) with ASV sequences named by your preferred ASV identifier. Both the ASV sequence and identifier will be merged with the output dataframe. If NULL, ASV-identifying data are excluded in the output dataframe.

return.conf

If TRUE, returns a list where the first element is your formatted taxonomy table and the second element is a dataframe of bootstrap confidence values. If FALSE, your formatted taxonomy table is returned as a dataframe.

Value

a dataframe formatted for use with taxmapper and/or ensembleTax

Details

For consistency with dada2's assignTaxonomy function, when used with Silva, RDP, or GreenGenes it subsamples the ranks c("domain", "phylum", "class", "order", "family", "genus"). Set db = NULL and supply ranks for databases that aren't directly supported. If a rubric is supplied with ASV-identifying meta-data (this is highly recommended), the output taxonomy table is sorted by the (first returned column of) ASV-identifying data.

Examples

Run this code

# NOT RUN {
data("bayes.sample")
data("rubric.sample")
head(bayes.sample)
head(rubric.sample)
df <- bayestax2df(tt = bayes.sample, db = "pr2", boot = 0, rubric = NULL,
return.conf = FALSE)
head(df)
df <- bayestax2df(tt = bayes.sample, db = "pr2", boot = 0,
rubric = rubric.sample, return.conf = FALSE)
head(df)
df <- bayestax2df(tt = bayes.sample, db = "pr2", boot = 60,
rubric = rubric.sample, return.conf = FALSE)
head(df)
df <- bayestax2df(tt = bayes.sample, db = "pr2", boot = 60,
rubric = rubric.sample, return.conf = TRUE)
head(df)

# }