isobar-preprocessing: IBSpectra preprocessing

In isobar: Analysis and quantitation of isobarically tagged MSMS proteomics data

Description

Preprocessing is a necessary step prior to analysis of data. In a sequential order, it is often neccassary to correct isotope impurities, to normalize, and subtract additive noise.

Isotope impurity correction

correctIsotopeImpurities(x):

Returns impurity corrected IBSpectra object by solving a linear system of equations. See also isotopeImpurities.

Normalization

normalize(x,f=median,target="intensity",exclude.protein=NULL, use.protein=NULL,f.doapply=TRUE,log=TRUE,channels=NULL,na.rm=FALSE):

Normalizes the intensities for multiplicative errors. Those changes are most likely produced by pipetting errors, and different hybridization efficencies, but can also be due to biological reasons. By default, tag intensities are multiplied by a factor so that the median intensity is equal across tags.

f:

f is applied to each column, unless f.doapply is FALSE. Then f is supposed to compute column-wise statistics of the matrix of intensities. E.g. colSums and colMeans.

target:

One of "intensity" and "ratio".

exclude.proteins

Spectra of peptides which might come from these proteins are excluded. Use for example for contaminants and proteins depleted in the experiment.

use.protein:

If specified, only spectra coming from this protein are used. Use when a protein is spiked-in as normalization control.

f.isglobal:

If true, f is applied on each column. If false, f is supposed to compute column-wise statistics of the matrix of intensities. E.g. colSums and colMeans.

log:

Used when target=ratio.

Substract additive noise

subtractAdditiveNoise(x,method="quantile",shared=TRUE,prob=0.01):

method

'quantile' method is supported for now. It take's the prob (0.01) quantile to estimate the noise level. This value is subtracted from all intensities, and all remaining intensities have to be at least that value.

prob

See 'method'.

shared

If channels are assumed similar in intensity and hence a shared noise level is reasonable. If not, then one level per channel is necessary.

Exclusion of proteins

exclude(x,proteins.to.exclude):

Removes spectra which are assigned to proteins in protein.to.exclude from the object. This can be useful to remove contaminants. It create a new grouping based on the data which is left.

proteins.to.exclude

Proteins to exclude.

Author(s)

Florian P. Breitwieser, Jacques Colinge

See Also

ProteinGroup, IBSpectra, isobar-analysis, isobar-plots

Examples

data(ibspiked_set1)
maplot(ibspiked_set1,main="IBSpiked, not normalized")
maplot(normalize(ibspiked_set1),main="IBSpiked, normalized")

isobar documentation built on Nov. 8, 2020, 7:48 p.m.