checkClass {LipidMS}

Search of class fragments to confirm the lipid class.

Description

Search of characteristic fragments that confirm a given lipid class.

Usage

checkClass(candidates, coelfrags, clfrags, ftype, clrequisites, ppm = 10, dbs)

Arguments

candidates

output of findCandidates function.

coelfrags

list of peaks coeluting with each candidate. Output of coelutingFrags.

clfrags

vector containing the expected fragments for a given lipid class. See details.

ftype

character vector indicating the type of fragments in clfrags. It can be: "F" (fragment), "NL" (neutral loss) or "BB" (building block). See details.

clrequisites

logical vector indicating if each class fragment is required or not. If none of the fragment is required, at least one of them must be present within the coeluting fragments. If the presence of any fragment excludes the class, it can be specified by using "excluding".

ppm

m/z tolerance in ppm.

dbs

list of data bases required for the annotation. By default, dbs contains the required data frames based on the default fragmentation rules. If these rules are modified, dbs may need to be changed. If data bases have been customized using createLipidDB, they also have to be modified here. It is employed when some fragment belongs to "BB" ftype.

Details

clfrags, ftype and clrequisites will indicate the rules to confirm a lipid class. All three arguments must have the same length.

This function allows three different types of fragments: fragments with a specific m/z as for example 227.0326 for PG in negative mode, which needs to be defined as clfrags = c(227.0326) and ftype = c("F"); neutral losses such as the head group of some PL (i.e. NL of 74.0359 in PG in negative mode), which will be defined as clfrags = c(74.0359) and ftype = c("NL"); or building blocks resulting from the loss of some groups, as for example, PA as M-H resulting from the loss of the head group (glycerol) in PG in ESI-, which will be defined as clfrags = c("pa_M-H") and ftype = c("BB"). The last two options could define the same fragments. In this case just one of them would be necessary.

When using the third type of fragment ("BB"), the building block will be specified in lower case (i.e. pa, dg, lysopa, mg, etc.) and the adduct will be given as it appears in the adductsTable, both separated by "_". Names for the building blocks are the ones used for the LipidMS databases without the "db" at the end.

In case the presence of a fragment indicates that the candidate does not belong to the lipid class (i.e. loss of CH3 in PE, which corresponds to a PC actually), this will be specified by using clrequisites = c("excluding").

Value

List with 2 elements: a matrix with logical values (presence/absense) of each expected fragment (columns) for each candidate (rows), and a logical vector with the confirmation of the lipid class for each candidate.

Author(s)

M Isabel Alcoriza-Balaguer <[e-mail]>