metID.chemSim: tanimoto chemical similarity first network neighbours

In WMBEdmands/CompMS2miner: an automatable metabolite identification, visualization and data-sharing R package for high-resolution LC-MS datasets

Description

tanimoto chemical similarity first network neighbours

Usage

metID.chemSim(object, ...)

Arguments

object	a "compMS2" class object.
autoPossId	logical if TRUE the function will automatically add the name of the top annotation based on mean maximum 1st neighbour chemical similarity above the minimum chemical similarity score (default = FALSE). Caution if TRUE this will overwrite any existing possible_identities in the "metID comments" table. This functionality is intended as an automatic annotation identification tool prior to thorough examination of the data in compMS2Explorer. The intention is that automatic annotations can be used in the metID.rtPred retention prediction function as part of a seamless first-pass workflow.
minSimScore	numeric must be values between 0-1 minimum tanimoto chemical similarity score (default = 0.8). Any mean maximum 1st neighbour chemical similarity scores will be considered for automatic possible identity addition to the "metID comments" table, the annotation with the highest mean maximum 1st neighbour chemical similarity score will then be automatically added to the "metID comments" table in the compMS2Explorer.
possContam	numeric how many times does a possible annotation have to appear in the automatically generated possible annotations for it to be considered a contaminant and therefore not added to the "metID comment" table (default = 3, i.e. if a database name appears more than 3 times in the automatic annotation table it will be removed).
bitsChemFP	numeric values between 1024-4096 number of most frequent atom-pairs in the DrugBank database see the ChemmineR function desc2fp for more details (default = 1024).
minEdges	numeric minimum number of edges (i.e. connected/adjacent nodes/spectra) to consider a node/spectrum for chemical similarity (default=2). This filtration is performed after removal of isobaric spectra. For example nodes with only one edge/adjacent node are more likely to produce false-positive annotations. For more robust chemical similarity scoring consider increasing this number.
verbose	logical if TRUE display progress bars.

Details

this function can only be utilized after running metID.corrNetwork and/or metID.specSimNetwork. The purpose of this function is to provide first-pass automatic metabolite annotation. The tanimoto chemical similarity score is first calculated from a 1024-4096 bit chemical fingerprint for every best annotation SMILES code. For annotations of each composite spectrum the maximum chemical similarity score with any first neighbours (either by correlation and/or spectral similarity) are identified and the weighted arithmetic mean maximum chemical similarity score of 1st neighbours calculated. The mean is weighted based on the mean spectral similarity and/or correlation coefficient value for an edge pair (i.e. if two spectra are connected by both spectral similarity and correlation then a mean value of the two will be calculated and used as the weight). This is to ensure that the more similar or highly correlated two composite spectra are the higher the contribution to the maximum chemical similarity score. A new column "MMNNCSS" is added to the best annotation tables for any composite spectra with at least one composite spectrum network neighbour. Furthermore, the best Annotation table is sorted according to this new likely annotation score, this give the user a rapid means to establish a likely annotation based on chemical similarity with neighbouring node annotations.

Optionally, the top annotations for a composite spectrum can be automatically added to the "metID comments" table in the compMS2Explorer application. If the argument autoPossId is TRUE (default = FALSE) the function will automatically add the name of the top annotation based on mean maximum 1st neighbour chemical similarity above the minimum chemical similarity score (argument minSimScore, default = 0.8).

Examples

library(compMS2Miner)
compMS2Example <- metID(compMS2Example, 'chemSim', minSimScore=0.8, 
                        autoPossId=TRUE)

WMBEdmands/CompMS2miner documentation built on May 9, 2019, 10:02 p.m.