R: export.class.table

export.class.table

R Documentation

export.class.table

Description

Builds a matrix with the probability for all mass to candidate compounds assignments, by averaging the number of assignments obtained by the gibbs sampler algorithm or ordering the compound candidates with the likelihood matrix.

Usage

  export.class.table(gibbsL = NULL, reactionM,
    molIon = NULL, probM = NULL, html = FALSE,
    filename = "test", burnIn = 3000, linkPattern = "kegg",
    m.test = "none", class1 = NULL, class2 = NULL,
    norm = FALSE, DB, prob = "count")

Arguments

`gibbsL`	a list of attributions and probabilities from gibbs.samp function.
`reactionM`	data.frame with compound annotation information.
`molIon`	non redundant ion annotation.
`probM`	optionally to gibbsL, a matrix of likelihoods.
`html`	logical, indicating whether a html file should be generated. This parameter uses the raw data to plot EICs and may be time consuming.
`filename`	html file name, the default is "test".
`burnIn`	how many samples of the gibbs sampler should be discarded.
`linkPattern`	which pattern should be linked to compound id, for now we have implemented "kegg", "pubchem" and "chebi" patterns.
`m.test`	statistical test to compare mean differences. This option is only available to single acquisition mode analysis, with options "t.test" and "anova".
`class1`	if the m.test is "t.test" first class to compare in the test, according with xcmsSet phenoData.
`class2`	if the m.test is "t.test" second class to compare in the test, according with xcmsSet phenoData.
`norm`	logical, if TRUE performs median normalization from (Anal. Chem. 2011, 83, 5864-5872).
`DB`	data.frame table used to search compounds, with the field name to be incorporated in the html table.
`prob`	how to calculate the probability to attribute a mass to a compound. Default is "count", which divide the number of times each identity was was attributed by the number of samples. Optionally the user could choose to use the mean of the probabilities of the identity, "mean".

Value

A list with a matrix "classTable" with attributions and probabilities and indexes of selected masses from xcms peak table.