dynamicNoiseFilter: Dynamic Noise filtration

In WMBEdmands/compMS2Miner: an automatable metabolite identification, visualization and data-sharing R package for high-resolution LC-MS datasets

Description

Dynamic Noise filtration

Usage

dynamicNoiseFilter(spectrum.df = NULL, DNF = 2, minPeaks = 5,
  maxPeaks = 20, minInt = 100)

Arguments

spectrum.df	a dataframe or matrix with two columns: 1. Mass/ Mass-to-charge ratio 2. Intensity
DNF	dynamic noise filter minimum signal to noise threshold (default = 2), calculated as the ratio between the linear model predicted intensity value and the actual intensity.
minPeaks	minimum number of signal peaks following dynamic noise filtration (default = 5).
maxPeaks	maximum number of signal peaks the function will continue until both the minimum DNF signal to noise ratio is exceeding and the number of peaks is lower than the maximum (default = 5).

Details

Dynamic noise filter adapted from the method described in Xu H. and Frietas M. "A Dynamic Noise Level Algorithm for Spectral Screening of Peptide MS/MS Spectra" 2010 BMC Bioinformatics.

The function iteratively calculates linear models starting from the median value of the lower half of all intensities in the spectrum.df. The linear model is used to predict the next peak intensity and ratio is calculated between the predicted and actual intensity value.

Assuming that all preceeding intensities included in the linear model are noise, the signal to noise ratio between the predicted and actual values should exceed the minimum signal to noise ratio (default DNF = 2).

The function continues until either the DNF value minimum has been exceeded and is also below the maxPeaks or maximum number of peaks value. As the function must necessarily calculate potentially hundreds of linear models the RcppEigen package is used to increase the speed of computation.

Value

a list containing 3 objects:

1. Above.noise The dynamic noise filtered matrix/ dataframe

2. metaData a dataframe with the following column names: 1. Noise.level the noise level determined by the dynamic noise filter function. 2. IntCompSpec Total intensity composite spectrum. 3. TotalIntSNR Sparse ion signal to noise ratio (mean intensity/ stdev intensity) 4. nPeaks number of peaks in composite spectrum

3. aboveMinPeaks Logical are the number of signals above the minimum level

WMBEdmands/compMS2Miner documentation built on May 9, 2019, 10:04 p.m.