table_annotate: Annotate LCMS features
In brgordon17/coralclass: Metabolomics Classification of Stressed Coral From Heron Island 2011

Description Usage Arguments Details Value Note Author(s) See Also

table_annotate() compares the 20 most important variables identified in the PLS-DA and Random Forests modelling of the LCMS data with the monoisotopic mass of those in coralmz.

1	table_annotate(ppm = 50, save.impvars = FALSE, save.matches = FALSE)

`ppm`	Numeric mass error to use for cross referencing
`save.impvars`	Logical indicating if the table of the 20 most important variables for each model should be saved to `./inst/extdata/`
`save.matches`	Logical indicating if the table of cross referencing matches should be save to `./tables/`

table_annotate() Loads the LCMS PLS-DA and LCMS RF models and creates a table (list$imp_vars) of the 20 most important variables according to varImp. The cross referencing relies on the user defined mass error. Three variables are created to define the upper and lower bounds of the mass error:

$mz_neutral: The neutral mass of the ion calculated by subtracting the monoisotopic mass of the hydrogen ion (1.007276 Da)
$mz_low: Equal to mz_neutral - the user defined ppm error
$mz_high: Equal to mz_neutral + the user defined ppm error

The table of the 20 most important variables and their mass error can be optionally saved to ./inst/extdata/ as a .rds file. Matches with compounds in the coralmz package, if any, are retrieved and can be optionally saved to ./tables/ as a comma separated .txt file.

Returns a list containing the important and matched variables

Commas in the references variable are replaced with semi colons to avoid conflicts with downstream use of the .csv file. Although this function is exported, table_crossref() was not intended to be used outside of this package.

Benjamin R. Gordon

varImp coralmz

brgordon17/coralclass documentation built on June 15, 2020, 9:21 p.m.