R Graphical Manual

Browse All

Last data update: 2014.03.03

R: ASCA Removal of Systematic Noise on Seq data

ARSyNseq

R Documentation

ASCA Removal of Systematic Noise on Seq data

Description

ARSyNseq filters the noise associated to identified or not identified batch effects considering the experimental design and applying Principal Component Analysis (PCA) to the ANOVA parameters and residuals.

Usage

ARSyNseq(data, factor = NULL, batch = FALSE, norm = "rpkm", logtransf = FALSE, Variability = 0.75, beta = 2)

Arguments

`data`	A Biobase's eSet object created with the `readData` function.
`factor`	Name of the factor (as it was given to the readData function) to be used in the ARSyN model (e.g. the factor containing the batch information). When it is NULL, all the factors are considered.
`batch`	TRUE to indicate that the factor argument indicates the batch information. In this case, the factor argument must be used to specify the names of the onlu factor containing the information of the batch.
`norm`	Type of normalization to be used. One of “rpkm” (default), “uqua”, “tmm” or “n” (if data are already normalized). If length was provided through the readData function, it will be considered for the normalization (except for “n”). Please note that if a normalization method if used, the arguments lc and k are set to 1 and 0 respectively.
`logtransf`	If FALSE, a log-transformation will be applied on the data before computing ARSyN model to improve the results of PCA on count data.
`Variability`	Parameter for Principal Componentents (PCs) selection of the ANOVA models effects. This is the desired proportion of variability explained for the PC of the main effects (time and experimental group). Variability=0.75 by default.
`beta`	Parameter for PCs selection of the residual model. Components selected will be those that explain more than beta times the average component variability computed as the total data variability divided by the rank of the matrix associated to the factor. Default beta=2.

Details

When batch is identified with one of the factors described in the argument factor of the data object, ARSyNseq estimates this effect and removes it by estimating the main PCs of the ANOVA effects associated. Selected PCs will be those that explain more than the variability proportion specified in Variability.

When batch is not identified, the model estimates the effects associated to each factor of interest and analyses if there exists systematic noise in the residuals. If there is batch effect, it will be identified with the main PCs of these residuals. Selected PCs will be those that explain more than beta times the average component variability.

Value

The Biobase's eSet object created with the readData function that was given as input but replacing the expression data with the filtered expression data matrix.

Author(s)

Maria Jose Nueda, mj.nueda@ua.es

References

Nueda, M.J.; Ferrer, A. and Conesa, A. (2012) ARSyN: a method for the identification and removal of systematic noise in multifactorial time-course microarray experiments. Biostatistics 13(3), 553-566.

Examples


# Generating an artificial batch effect from Marioni's data
data(Marioni)
set.seed(123)
mycounts2 = mycounts
mycounts2[,1:4] = mycounts2[,1:4] + runif(nrow(mycounts2)*4, 3, 5)
myfactors = data.frame(myfactors, "batch" = c(rep(1,4), rep(2,6)))
mydata2 = readData(mycounts2, factors = myfactors)

# Exploring batch effect with PCA
myPCA = dat(mydata2, type = "PCA")
par(mfrow = c(1,2))
explo.plot(myPCA, factor = "Tissue")
explo.plot(myPCA, factor = "batch")

# Removing batch effect when the batch is identified for each sample and exploring results with PCA
mydata2corr1 = ARSyNseq(mydata2, factor = "batch", batch = TRUE, norm = "rpkm",  logtransf = FALSE)
myPCA = dat(mydata2corr1, type = "PCA")
par(mfrow = c(1,2))
explo.plot(myPCA, factor = "Tissue")
explo.plot(myPCA, factor = "batch")

# If we consider that exist a batch but it is not identified (we do not know the batch information):
mydata2corr2 = ARSyNseq(mydata2, factor = "Tissue", batch = FALSE, norm = "rpkm",  logtransf = FALSE)
myPCA = dat(mydata2corr2, type = "PCA")
par(mfrow = c(1,2))
explo.plot(myPCA, factor = "Tissue")
explo.plot(myPCA, factor = "batch")

Results


R version 3.3.1 (2016-06-21) -- "Bug in Your Hair"
Copyright (C) 2016 The R Foundation for Statistical Computing
Platform: x86_64-pc-linux-gnu (64-bit)

R is free software and comes with ABSOLUTELY NO WARRANTY.
You are welcome to redistribute it under certain conditions.
Type 'license()' or 'licence()' for distribution details.

R is a collaborative project with many contributors.
Type 'contributors()' for more information and
'citation()' on how to cite R or R packages in publications.

Type 'demo()' for some demos, 'help()' for on-line help, or
'help.start()' for an HTML browser interface to help.
Type 'q()' to quit R.

> library(NOISeq)
Loading required package: Biobase
Loading required package: BiocGenerics
Loading required package: parallel

Attaching package: 'BiocGenerics'

The following objects are masked from 'package:parallel':

    clusterApply, clusterApplyLB, clusterCall, clusterEvalQ,
    clusterExport, clusterMap, parApply, parCapply, parLapply,
    parLapplyLB, parRapply, parSapply, parSapplyLB

The following objects are masked from 'package:stats':

    IQR, mad, xtabs

The following objects are masked from 'package:base':

    Filter, Find, Map, Position, Reduce, anyDuplicated, append,
    as.data.frame, cbind, colnames, do.call, duplicated, eval, evalq,
    get, grep, grepl, intersect, is.unsorted, lapply, lengths, mapply,
    match, mget, order, paste, pmax, pmax.int, pmin, pmin.int, rank,
    rbind, rownames, sapply, setdiff, sort, table, tapply, union,
    unique, unsplit

Welcome to Bioconductor

    Vignettes contain introductory material; view with
    'browseVignettes()'. To cite Bioconductor, see
    'citation("Biobase")', and for packages 'citation("pkgname")'.

Loading required package: splines
Loading required package: Matrix
> png(filename="/home/ddbj/snapshot/RGM3/R_BC/result/NOISeq/ARSyNSeq.Rd_%03d_medium.png", width=480, height=480)
> ### Name: ARSyNseq
> ### Title: ASCA Removal of Systematic Noise on Seq data
> ### Aliases: ARSyNseq arsynseq
> ### Keywords: ASCA, ANOVA, PCA, batch
> 
> ### ** Examples
> 
> 
> # Generating an artificial batch effect from Marioni's data
> data(Marioni)
> set.seed(123)
> mycounts2 = mycounts
> mycounts2[,1:4] = mycounts2[,1:4] + runif(nrow(mycounts2)*4, 3, 5)
> myfactors = data.frame(myfactors, "batch" = c(rep(1,4), rep(2,6)))
> mydata2 = readData(mycounts2, factors = myfactors)
> 
> # Exploring batch effect with PCA
> myPCA = dat(mydata2, type = "PCA")
> par(mfrow = c(1,2))
> explo.plot(myPCA, factor = "Tissue")
> explo.plot(myPCA, factor = "batch")
> 
> # Removing batch effect when the batch is identified for each sample and exploring results with PCA
> mydata2corr1 = ARSyNseq(mydata2, factor = "batch", batch = TRUE, norm = "rpkm",  logtransf = FALSE)
> myPCA = dat(mydata2corr1, type = "PCA")
> par(mfrow = c(1,2))
> explo.plot(myPCA, factor = "Tissue")
> explo.plot(myPCA, factor = "batch")
> 
> # If we consider that exist a batch but it is not identified (we do not know the batch information):
> mydata2corr2 = ARSyNseq(mydata2, factor = "Tissue", batch = FALSE, norm = "rpkm",  logtransf = FALSE)
> myPCA = dat(mydata2corr2, type = "PCA")
> par(mfrow = c(1,2))
> explo.plot(myPCA, factor = "Tissue")
> explo.plot(myPCA, factor = "batch")
> 
> 
> 
> 
> 
> dev.off()
null device 
          1 
>