NanoStringQCPro (version 1.4.0)

nSolverBackground,RccSet-method: nSolver Analysis Software background estimation

Description

Calculates initial probe- and lane-specific background estimates using an algorithm that mimics the implementation in NanoString's nSolver Analysis Software (see details below for the exact algorithm).

Usage

"nSolverBackground"(rccSet, stringency = 1, shrink = TRUE, w1 = 2.18, inputMatrix = c("posCtrlData", "exprs"))

Arguments

rccSet

NanoString RccSet object

stringency

Multiplier with which to adjust final values.

shrink

Boolean specifying if probe-specific estimates should be shrunken towards their global mean.

w1

Shrink weight "w1".

inputMatrix

Name of the matrix in the RccSet's assayData to use as input for calculating background estimates (one of "exprs" or "posCtrlData"). If posCtrlData is specified but not present in the assayData, an error will be generated.

Value

A matrix containing lane- and probe-specific background estimates.

Details

The mean values for each blank lane (not including positive control probes) are computed from the original data, and a vector of probe-specific background is established by taking the rowMeans of the blank measurements for each probe after subtracting out these values. If shrink=TRUE, the vector is adjusted via the following formula (where 'probe.bg' represents the vector):

w2 <- 1/length(blanks)
probe.bg <- (w1*probe.bg + w2*mean(probe.bg)) / (w1 + w2)

This probe-specific background is further adjusted by subtracting the mean of its values for the negative control probes. A lane-specific "affinity" is calculated for all lanes in the original data by taking the colMeans of the negative control probe values in the original data, and background estimates for each probe and lane in the original data are computed by summing the corresponding probe-specific background and lane-specific affinity. Any resulting values less than zero are set to zero, and the last step before returning these values is to multiply them by the given stringency.

See Also

getBackground, subtractBackground