Skip to contents

Finds the LAPALA

Source: R/missingValuesImputation_ProteinLevel.R
mv_imputation_protein.Rd

Methods available are:

  • wrapper.impute.detQuant(): This method is a wrapper of the function impute.detQuant() for objects of class MSnSet

  • wrapper.impute.KNN(): Can impute only POV missing values. This method is a wrapper for objects of class QFeatures and imputes missing values with a fixed value. This function imputes the missing values condition by condition.

  • wrapper.impute.slsa(): Imputation of peptides having no values in a biological condition. This method is a wrapper to the function impute.slsa() of the package imp4p adapted to an object of class MSnSet.

  • wrapper.impute.fixedValue(): This method is a wrapper to objects of class MSnSet and imputes missing values with a fixed value.

  • wrapper.impute.pa(): Imputation of peptides having no values in a biological condition. This method is a wrapper to the function impute.pa of the package imp4p adapted to an object of class MSnSet.

Usage

findMECBlock(obj, grp)

reIntroduceMEC(obj, grp, MECIndex)

wrapper.impute.KNN(obj = NULL, grp, K)

wrapper.impute.fixedValue(obj, grp, fixVal = 0, na.type)

wrapper.impute.pa(obj = NULL, grp, q.min = 0.025)

wrapper.impute.detQuant(obj, qval = 0.025, factor = 1, na.type)

getQuantile4Imp(qdata, qval = 0.025, factor = 1)

wrapper.impute.slsa(obj = NULL, design = NULL)

Arguments

obj

An object of class QFeatures.

grp

xxx

MECIndex

A data.frame that contains index of MEC (see findMECBlock)

K

the number of neighbors.

fixVal

A float.

na.type

A string which indicates the type of missing values to impute. Available values are: NA (for both POV and MEC), POV, MEC.

q.min

Same as the function impute.pa() in the package imp4p

qval

An expression set containing quantitative values of various replicates

factor

A scaling factor to multiply the imputation value with

qdata

xxx

design

xxx

Value

A data.frame containing the indexes of LAPALA

A list of two vectors, respectively containing the imputation values and the rescaled imputation values

Utilities functions

  • findMECBlock(): xxx

  • reIntroduceMEC(): xxx

  • getQuantile4Imp(): Quantile imputation value definition. This method returns the q-th quantile of each column of an expression set, up to a scaling factor

Author

Samuel Wieczorek

Examples

data(subR25prot)
obj <- subR25prot[[2]]
grp <- design.qf(subR25prot)$Condition
lapala <- findMECBlock(obj, grp)
na.type = c("Missing POV", "Missing MEC")
obj.imp.pov <- wrapper.impute.detQuant(obj, na.type = na.type)
obj.imp.pov <- reIntroduceMEC(obj, grp, lapala)

obj.imp.pov <- wrapper.impute.KNN(obj, grp, 3)
#> Warning: 2 rows with more than 99 % entries missing;
#>  mean imputation used for these rows
#> Warning: 2 rows with more than 99 % entries missing;
#>  mean imputation used for these rows

obj.imp.pov <- wrapper.impute.fixedValue(obj, grp, 0.001, na.type = "Missing POV")
obj.imp.mec <- wrapper.impute.fixedValue(obj, grp, 0.001, na.type = "Missing MEC")
obj.imp.na <- wrapper.impute.fixedValue(
obj, grp, 0.001, 
na.type = c("Missing MEC", "Missing POV"))

obj.imp.pov <- wrapper.impute.pa(obj, grp)

qdata <- SummarizedExperiment::assay(obj)
quant <- getQuantile4Imp(qdata)