RPPanalyzer (version 1.2.2)

calcSdc: Calculates the concentration of serial diluted samples

Description

Calculates the protein concentration of a serial diluted sample stored in an RPPA data list using the serial dilution curve algorithm published by Zhang et.al, Bioinformatics 2009.

Usage

calcSdc(x,sample.id=c("sample","sample.n"),
		sel=c("measurement","control"), dilution="dilution",
		D0=2,sensible.min=5, sensible.max=1.e9,minimal.err=5,
		plot=T, r=1.2)

Arguments

x

RPPA data list with replicates aggregated with median

sample.id

Attributes to identify the samples

sel

The sample type that should be calculated. Has to be "measurements","control", "neg_control",or "blank".

dilution

Name of the column in the feature data matrix describing the dilution steps of the samples.

D0

Dilution factor.

sensible.min

Signals below this value are marked as undetected

sensible.max

Signals above the value are marked as saturated

minimal.err

Minimal valid estimate for the background noise

plot

Logical. If true, model fits are plotted

r

Constant factor used to determine the confidence interval for the saturation limit $M$ and the background noise $a$, shoul be $>1$. Can be lower if accuracy of signals is improved.

Value

• expression
  matrix with expression values
• error
  matrix with error values
• arraydescription
  data frame with feature data
• sampledescription
  data frame with pheno data

Details

The method of Zhang et. al doesn't fit the dose response curve but a derive model describing the functional relationship between the signals of two consecutive dilution steps. Since this new model does not contain the protein concentration anymore all spots of one array can be used for the fit, allowing a much more robust estimation of the underlying paramters.

References

Zhang et. al, Bioinformatics 2009,Serial dilution curve: a new method for analysis of reverse phase protein array data

Examples

library(RPPanalyzer)
    data(ser.dil.samples)

    ser.dil_median <- sample.median(ser.dil.samples)
    predicted.data <- calcSdc(ser.dil_median,D0=2,sel=c("measurement"), dilution="dilution")