computeSimilarity: Compute Spectral Similarity

In pmbrophy/mzDataTable: Import and convert raw mass spectrometry data formats

Description

Compute Spectral Similarity

Usage

computeSimilarity(
  mzDt,
  ppmTol,
  isCentroid,
  intensityMin = 0,
  normalize = TRUE,
  normalization_method = "sqrt",
  similarity_method
)

Arguments

mzDt	a data.table imported by mzML2dataTable() with columns 'seqNum', 'mz', 'intensity'
ppmTol	the mass accuracy of the instrument
isCentroid	Logical. Is the data centered?
intensityMin	Optional parameter. Ions with intensity < intensityMin will be removed. Default = 0 allowing all ions into algorighm.
normalize	Optional parameter. Should the intensity be normalized before any processing is done? Default = TRUE.
normalization_method	Optional parameter. Select normalization method: "maxPeak", "sqrt", "sum". See '.normalizeEachSpectrum_dt' for details.
similarity_method	Select similarity metric to be used. "dotProd", "specContrast", "specCor", "brayCurtis", "euclidean".

Value

a 3 column matrix containing the similarity score and seqNum of each compared spectrum

Examples

dt <- mzML2dataTable(path = msdata::proteomics(full.names = TRUE)[3])
simMat <- computeSimilarity(mzDt = dt[msLevel == 2 & seqNum %between% c(1, 10)], 
                            ppmTol = 100, 
                            isCentroid = TRUE, 
                            intensityMin = 0.01, 
                            normalize = TRUE, 
                            normalization_method = "sqrt", 
                            similarity_method = "specContrast")

squareMat <- similarity2SquareMatrix(simMat)
image(squareMat)

pmbrophy/mzDataTable documentation built on June 6, 2020, 7:43 a.m.