PreProcessPam: Pre-processing Cell Line specific data for a given Peptide

In pamgene/PamGeneMixed: Preprocessing and Modeling Kinase Activity Profiles in PamChip Data

Description

This function removes Cell Line specific extreme observations, profiles and negative intensity measurements of a given peptide. Negative intensity measurements within a profile are replaced by half of the positive minimum intensity measurement of that profile. Extreme profiles have been removed via residuals based on one-way ANOVA (Response for the ANOVA is AUC: Average(residuals) plus or minus 2 times SE(residuals)). Extreme observations within a profile were deleted if they differ more than plus or minus 2 times standard deviations of the profile-specific mean.

Usage

PreProcessPam(p=108, cel =4, d=2, PamS, plotting=TRUE )

Arguments

p	Peptide to be considered. And should be supplied as numeric value that corresponds to the peptide index.
cel	Cell Line index.
d	Flexibility for the extreme observations to be removed. Eg. 2 or 3 times standard deviations. Default is 2.
PamS	A PamChipData data frame.
plotting	A Boolean parameter. Default is TRUE and observed and filtered data are visualized for a given cell line of a peptide.

Value

A list of objects of class CleanedPeptide.

Author(s)

Pushpike Thilakarathne, Ziv Shkedy and Dan Lin

See Also

PreProcessAllPeptides

Examples

  data(PamChipData)


#-----------
PreProcessPam(p   =20, cel =4,   d   =2, PamS=PamChipData, plotting=T )

#-----------
PreProcessPam(p   =6, cel =4,   d   =2, PamS=PamChipData, plotting=T )

#-----------
PreProcessPam(p   =78, cel =14,   d   =2, PamS=PamChipData, plotting=T)

#-----------
PreProcessPam(p   =143, cel =2,   d   =2, 
PamS=PamChipData[PamChipData[,c("ResState")]=="R",],  plotting=T )

#----------- Sequentially visualize the preprocessing steps -----------
for (i in 1:148){
    for (cc in 1:20){
    par(ask=T)
    PreProcessPam(p   =i, cel =cc,   d   =2, PamS=PamChipData, plotting=T )
    }
}

    

pamgene/PamGeneMixed documentation built on Dec. 31, 2020, 1:13 a.m.