beadarrayFilter: An R Package to Filter Beads
Anyiawung Chiara Forcheh
Interuniversity Institute for Biostatistics and statistical Bioinformatics,Geert Verbeke and Lieven Clement
Interuniversity Institute for Biostatistics and statistical Bioinformatics, Katholieke Universiteit Leuven,Dan Lin and Ziv Shkedy
Interuniversity Institute for Biostatistics and statistical Bioinformatics, Universiteit Hasselt,Adetayo Kasim
Wolfson Research Institute, Durham University,Willem Talloen and Hinrich W.H.
Janssen Pharmaceutica N.V.,2013-06-02
Abstract
Microarrays enable the expression levels of thousands of genes to be measured simultaneously. However, only a small fraction of these genes are expected to be expressed under different experimental conditions. Nowadays, filtering has been introduced as a step in the microarray pre-processing pipeline. Gene filtering aims at reducing the dimensionality of data by filtering redundant features prior to the actual statistical analysis. Previous filtering methods focus on the Affymetrix platform and can not be easily ported to the Illumina platform. As such, we developed a filtering method for Illumina bead arrays. We developed an R package, beadarrayFilter, to implement the latter method. In this paper, the main functions in the package are highlighted and using many examples, we illustrate how beadarrayFilter can be used to filter bead arrays.Introduction
Gene expression patterns are commonly assessed by microarrays that can measure thousands of genes simultaneously. However, in a typical microarray experiment, only a small fraction of the genes are informative which motivated the development of gene filtering methods. Gene filtering aims at reducing the dimensionality of data by filtering redundant features prior to the actual statistical analysis. This has been shown to improve differential expression analysis with Affymetrix microarrays (e.g., Talloen, D.-A. Clevert, S. Hochreiter, D. Amaratunga, L. Bijnens, S. Kass, and H. W. H. Göhlmann 2007; Calza, W. Raffelsberger, A. Ploner, J. Sahel, T. Leveillard, and Y. Pawitan 2007; Kasim, D. Lin, S. Van Sanden, D.-A. Clevert, L. Bijnens, H. Göhlmann, D. Amaratunga, S. Hochreiter, Z. Shkedy, and W. Talloen 2010).
Although different microarrays platforms share the same principle of
hybridizing DNA to a complementary probe, they differ considerably by
design. Unlike the Affymetrix microarrays which have sets of unique
probes targeting a particular gene (resulting in a probe set for a
targeted gene), Illumina microarrays have sets of identical probes.
Thus, the existing filtering methods can not be readily ported to the
Illumina platform. As a result, Forcheh,
G. Verbeke, A. Kasim, D. Lin, Z. Shkedy, W. Talloen, H. W. Göhlmann, and
L. Clement (2012) developed a filtering method for Illumina bead
arrays. Forcheh, G. Verbeke, A. Kasim, D. Lin,
Z. Shkedy, W. Talloen, H. W. Göhlmann, and L. Clement (2012)
equally showed that filtering improves the analysis of differential
expression. We provide the implementation of their method in the beadarrayFilter
R software package. The beadarrayFilter package can take a
normalized data frame or a normalized bead array ExpressionSetIllumina
object (obtained using the summarize
or readBeadSummaryData functions in the Bioconductor
package beadarray by Dunning, M. L. Smith, M. E. Ritchie, and
S. Tavaré 2007) or a normalized LumiBatch object as input and
returns a list containing a filtered data frame or a filtered bead array
ExpressionSetIllumina object or a filtered LumiBatch object,
respectively. The package can also process summarized and normalized
average intensities (eSet), their standard errors (seSet) and the number
of beads used for summarization (nSet) as input and returns a list of
components including the intra-cluster correlations (ICC), which can be
used to assess different filtering strategies.
The paper contains a brief background of the filtering methodology followed by the introduction of the beadarrayFilter package with illustrative examples.
Filtering Illumina bead types
Let \(Y_{kij}\) be the bead-level expression intensity of bead \(j\) in sample (array) \(i\) in treatment group \(k\), \(i = 1,\ldots, n_k\), \(j = 1,\ldots,m_{ki}\), \(k = 1,\ldots, K\), then, Forcheh, G. Verbeke, A. Kasim, D. Lin, Z. Shkedy, W. Talloen, H. W. Göhlmann, and L. Clement (2012) proposed the following filtering model:
\[Y_{kij} = \mu + b_i + \epsilon_{kij}, \label{eq:filter} (\#eq:filter) \]
where \(\mu\) represents the average expression for a bead type across all samples, \(b_i\) is the array specific random effect and \(\epsilon_{kij}\) are the measurement errors, both normally distributed with mean zero and variance \(\tau^2\) and \(\sigma^2\), respectively:
\(b_i \sim N(0,\tau^2)\), \(\epsilon_{kij} \sim N(0,\sigma^2)\).
The \(\tau^{2}\) captures the between-array variability while \(\sigma^2\) models the within-array variability. Model ([eq filter]) implies a common ICC, (Verbeke and G. Molenberghs 2000), given by
\[\rho = \frac{\tau^2}{\tau^2 + \sigma^2} , \label{eq:intracl} (\#eq:intracl) \]
which is the correlation among the replicate probes on the array.
An informative bead type is expected to have a relatively high value of \(\rho\) since all corresponding beads have the same sequence. As such, the between-array variability is the dominant variance component for informative bead types while the within-array variability is the largest variance component for the non-informative bead types.
For Illumina bead arrays, within-array variability is expected to vary across all arrays and treatment groups in the experiment. In this regard, Forcheh, G. Verbeke, A. Kasim, D. Lin, Z. Shkedy, W. Talloen, H. W. Göhlmann, and L. Clement (2012) included an array/treatment specific variance component to adjust the model for heteroscedasticity i.e., \(\epsilon_{kij} \sim N(0,\sigma^2_{ki})\). Hence, the ICC becomes bead/array/group specific,
\[\rho_{ki} = \frac{\tau^2}{\tau^2 + \sigma^2_{ki}, } \label{eq:intrac} (\#eq:intrac) \]
and \(\sum_{k=1}^K n_k\) ICCs are obtained for each bead type (more details can be found in Forcheh, G. Verbeke, A. Kasim, D. Lin, Z. Shkedy, W. Talloen, H. W. Göhlmann, and L. Clement 2012).
ICC based filtering of Affymetrix microarray data has been proposed in the literature (e.g., Talloen, D.-A. Clevert, S. Hochreiter, D. Amaratunga, L. Bijnens, S. Kass, and H. W. H. Göhlmann 2007; Kasim, D. Lin, S. Van Sanden, D.-A. Clevert, L. Bijnens, H. Göhlmann, D. Amaratunga, S. Hochreiter, Z. Shkedy, and W. Talloen 2010). Typically, an ICC cut-off \(\delta = 0.5\) has been used to declare a transcript informative. Forcheh, G. Verbeke, A. Kasim, D. Lin, Z. Shkedy, W. Talloen, H. W. Göhlmann, and L. Clement (2012) also relied on the ICC as a filtering criterion and proposed different thresholding schemes based on the five number summaries, i.e., a bead type can be called informative based on thresholding (1) the minimum ICC, (2) 25, (3) 50, (4) 75 quantiles or (5) the maximum ICC of all the arrays. Filtering is expected to become less stringent from strategy (1)–(5). These different thresholding strategies are implemented within the package.
beadarrayFilter package
The beadarrayFilter package can either take a normalized
ExpressionSetIllumina object, a normalized LumiBatch object, a
normalized data.frame or a normalized eSet, seSet and nSet as input and
returns a list. We refer the user to the documentation of the
Bioconductor packages beadarray and lumi for more
details on generating ExpressionSetIllumina objects or LumiBatch
objects. For each bead type, the ICCs can be summarized using the 5
number summary or user specified quantiles. The corresponding ICC
summaries are used for obtaining informative bead types. The package
contains two major functions, which we refer to as: (1) a low level
function iccFun and (2) a wrapper function
beadtypeFilter. Model fitting is done using a modified
version of the MLM.beadarray function of (Kim and J. Lin 2011). Details on the functions
can be obtained by using the help function in R
(?beadtypeFilter or ?iccFun).
Installation of beadarrayFilter
The beadarrayFilter package is available at CRAN, and can be
installed using
install.packages("beadarrayFilter").
beadtypeFilter function
The beadtypeFilter function is a wrapper function for
the iccFun function and is designed for users with a
primary interest in obtaining filtered bead types. This function takes a
normalized ExpressionSetIllumina object, a normalized LumiBatch object
or a normalized data.frame and returns the names of the informative bead
types. Optionally, the filtered ExpressionSetIllumina object or the
filtered data.frame can also be returned. The filtered
ExpressionSetIllumina object, or filtered LumiBatch object or the
filtered data.frame can then be used for the downstream analysis.
iccFun function
iccFun is a low level function. It is designed for users
who want to assess different filtering strategies. It takes a normalized
eSet, seSet and nSet and the bead types identification variable
(ProbeID), fits the filtering Model (@ref(eq:filter)), calculates the
ICC for each bead type on each array/treatment group, summaries the ICCs
at the specified quantiles, and returns the ICC summaries, the
within-array variances, the between-array variances as well as all ICCs.
The ICCs output can later be used for filtering or to assess different
filtering strategies. Note the information printed as you execute the
beadtypeFilter or iccFun functions:
\([1]\) "Number of converged transcripts: ... "
This indicates the number of transcripts for which the filtering model has already converged while "Now ... remaining..." tells the number of transcripts still to be processed.\([1]\) "Computing ICC for each array"
This message is printed when the function begins to calculate the ICC for each array once the filtering model has been fitted to all the transcripts.\([1]\) "Summarising the ICC at supplied quantile(s)"
This message indicates the last stage of the filtering function where the ICCs are summarized at the supplied quantiles (see Forcheh, G. Verbeke, A. Kasim, D. Lin, Z. Shkedy, W. Talloen, H. W. Göhlmann, and L. Clement 2012).
The emCDF function within the beadarrayFilter
package is used to plot the empirical cumulative density functions
(edcfs) for the different threshold strategies discussed above. It
processes the iccFun output and plots the empirical
cumulative density functions (ecdf) for the different threshold
strategies as discussed in Forcheh, G. Verbeke,
A. Kasim, D. Lin, Z. Shkedy, W. Talloen, H. W. Göhlmann, and L. Clement
(2012). It is expected that the filtering becomes more stringent
as the ICC threshold increases and/or as the the thresholded ICC
quantile decreases.
Informative bead types will have larger between-array variances as
compared to the within-array variances. The varianceplot
function takes the estimated between- and within-array variances from
the iccFun function as inputs and plots them. The variances
of noninformative bead types are plotted in blue while those of the
informative bead types are displayed in red.
Upon filtering the MLM.beadarray function can be used
for downstream analysis of single factor designs. For more details, see
(Kim and J. Lin 2011).
A summary of the functions within the beadarrayFilter package and their use is presented in Table 1.
| Function | Description |
|---|---|
beadtypeFilter() |
Fits the filtering model as in Forcheh, G. Verbeke, A. Kasim, D. Lin, Z. Shkedy, W. Talloen, H. W. Göhlmann, and L. Clement (2012), |
| computes the ICC and filters the bead types. | |
iccFun() |
Fits the filtering model as in Forcheh, G. Verbeke, A. Kasim, D. Lin, Z. Shkedy, W. Talloen, H. W. Göhlmann, and L. Clement (2012), and computes |
| the ICC which can later be used for filtering or to assess different | |
| filtering strategies | |
MLM.beadarray() |
Function to fit the filtering model or for the downstream analysis |
| of single factor experimental designs | |
emCDF() |
Plots the ecdf for the different threshold strategies |
varianceplot() |
Plots the the between-array and the within-array variances |
Examples
The data set exampleSummaryData from the Bioconductor
package beadarrayExampleData (Dunning,
M. L. Smith, M. E. Ritchie, and S. Tavaré 2007), is used to
illustrate filtering of ExpressionSetIllumina objects while the publicly
available spike-in data (Dunning, N. B. Morais,
A. Lynch, S. Tavaré, and M. Ritchie 2008) are used to process
data.frames. The \(\log_2\) summarized
expression intensities, obtained from Brain and Universal Human
Reference RNA (UHRR) samples in the data set
exampleSummaryData will be used. There are 12 arrays, 6 for
the Brain samples and 6 for the UHRR samples. The spike-in dataset
consists of 48 arrays and an array contained \(\sim\) 48,000 bead types (non-spikes) and
33 bead types (spikes) targeting bacterial and viral genes absent in the
mouse genome. We use the same subset of the spike-in data as in (Kim and J. Lin 2011). This dataset includes
34,654 bead types targeting annotated genes with Genebank IDs and the 33
spikes. The data can be downloaded from https://ephpublic.aecom.yu.edu/sites/rkim/Supplementary/files/KimLin2010.html
with the folder name “A normalized dataset for example analysis". We
also show how the downstream analysis can be performed upon filtering
using the spike-in data. Filtering of LumiBatch objects is illustrated
using the BeadStudio output used in the vignette”beadsummary” from the
beadarray package, which can be downloaded from http://www.switchtoi.com/datasets.ilmn.
beadtypeFilter function
This subsection shows how to use the beadtypeFilter
function to filter normalized ExpressionSetIllumina objects, LumiBatch
objects and data.frames. For an ExpressionSetIllumina, this is done
using the exampleSummaryData data set from the
beadarrayExampleData package.
# Normalize the log2 transformed data
> library("beadarrayFilter")
> data("exampleSummaryData", package = "beadarrayExampleData")
> exampleSummaryDataNorm <- normaliseIllumina(channel(exampleSummaryData, "G"),
+ method = "quantile", transform = "none")
# Filter the ExpressionSetIllumina
> iccResults <- beadtypeFilter(exampleSummaryDataNorm, Quantile = 1,
+ keepData = TRUE, delta = 0.5)By specifying iccQuant = 1 and delta = 0.5,
the bead types are filtered using the maximum ICC at a cutoff of 50%.
The output of the beadtypeFilter function can then be
observed as follows:
> head(iccResults$InformProbeNames)
[1] "ILMN_1802380" "ILMN_1736104" "ILMN_1792389" "ILMN_1705423" "ILMN_1697642"
[6] "ILMN_1788184"
> exprs(iccResults$informData)[1:6, 1:5]
4613710017_B 4613710052_B 4613710054_B 4616443079_B 4616443093_B
ILMN_1802380 8.216547 8.229713 8.097047 8.343822 8.249190
ILMN_1736104 5.317065 5.470957 5.054653 5.100678 5.446530
ILMN_1792389 6.725049 7.003632 6.783809 7.214921 7.257032
ILMN_1705423 5.496207 4.845898 5.394206 5.422772 5.479191
ILMN_1697642 7.977234 7.912246 7.668253 7.850134 7.758535
ILMN_1788184 5.291988 5.614500 5.565426 5.473346 5.573395
> head(fData(iccResults$informData))
ArrayAddressID IlluminaID Status
ILMN_1802380 10008 ILMN_1802380 regular
ILMN_1736104 10017 ILMN_1736104 regular
ILMN_1792389 10019 ILMN_1792389 regular
ILMN_1705423 10039 ILMN_1705423 regular
ILMN_1697642 10044 ILMN_1697642 regular
ILMN_1788184 10048 ILMN_1788184 regular
> dim(exampleSummaryDataNorm)
Features Samples Channels
49576 12 1
> dim(iccResults$informData)
Features Samples Channels
23419 12 123419 out of the 49576 bead types were declared informative using the maximum ICC at a cutoff point of 50%.
For a LumiBatch object, filtering is illustrated using the
non-normalized data, an output of BeadStudio used in the “beadsummary”
vignette from the beadarray package. The data file,
AsuragenMAQC_BeadStudioOutput.zip, can be downloaded from
http://www.switchtoi.com/datasets.ilmn. Once the file
has been downloaded, unzip its content to your R working directory.
> require(lumi)# Set the working directory to the directory where the unzipped data file was saved.
> setwd("C:/Multi_level_Illumina_feb2011/RPackageFinal/beadstudiooutputData")
# Read in the data using lumiR to obtain a LumiBatch object
> x.lumi <- lumiR("AsuragenMAQC-probe-raw.txt")
# Normalize the data without any further transformation step
> lumi.N <- lumiN(x.lumi, "rsn")
# Filter the LumiBatch
> iccResult <- beadtypeFilter(lumi.N, Quantile = 1, keepData = TRUE, delta = 0.5)By specifying iccQuant = 1 and delta = 0.5,
the bead types are filtered using the maximum ICC at a cutoff of
50%.
> dim(lumi.N)
Features Samples
48701 6
> dim(iccResult$informData)
Features Samples
1195 6Only 1195 of the 48701 bead types were declared informative using the maximum ICC at a cutoff of 50%. This may be due to the way the data was summarized and normalized. How the processing of bead array data affects bead types filtering is a topic of future research.
For a data.frame, the beadtypeFilter function is
illustrated using the Illumina spike-in data. Read the data from the
file location where the data had been downloaded, unzipped and
saved.
> filepath <- "C:/Multi_level_Illumina_feb2011/log2scale.normalized.txt"
> dt <- read.delim(filepath, header = TRUE, as.is = TRUE, row.names = NULL)[,-1]
> dt[1:6,1:5]
ProbeID X1377192001_A.AVG_Signal X1377192001_A.Detection.Pval
1 50014 6.150486 0.579207900
2 50017 6.616132 0.074257430
3 50019 8.164317 0.000000000
4 50020 7.414991 0.001856436
5 50022 5.804593 0.974628700
6 50025 6.412067 0.173267300
X1377192001_A.Avg_NBEADS X1377192001_A.BEAD_STDERR
1 27 0.09889349
2 40 0.05644992
3 25 0.06384269
4 27 0.07853792
5 38 0.08098911
6 28 0.08153830Note, that the data.frame supplied to the beadtypeFilter
function should contain the summarized intensities (eSet),
standard errors (seSet) and the number of beads used for
the summarization (nSet). When using a data frame, column
names should be conform to BeadStudio output, i.e., the column names for
eSet should end on "Signal", those for seSet on "STDERR" and the columns
corresponding to nSet should end on "NBEADS". It is preferable to use an
identification variable with a unique ID for each bead type. In the
spike-in data, the spikes all have the same TargetID, thus
the ProbeID is preferred. Similar to the
ExpressionSetIllumina example, the beadtypeFilter function
is used for filtering.
> iccResults <- beadtypeFilter(dt, Quantile = 0.5, keepData = TRUE, delta = 0.5)By specifying Quantile = 0.5, bead types are filtered
using the median ICC.
> head(iccResults$InformProbeNames)
[1] 50280 50440 70594 110138 110685 130402
> dim(dt)
[1] 34687 193
> dim(iccResults$informData)
[1] 238 193238 of the 34687 bead types were declared informative based on thresholding the median ICC at a cutoff of 50%. A large number of bead types have been filtered out. It should be noted that this is probably due to the artificial nature of the spike-in data and we would expect lesser bead types to be filtered out in real life data.
iccFun function
The examples in this section show how the iccFun
function can be used to process different data types, observe its output
and assess the filtering strategies.
Processing data using the iccFun function
Processing a data.frame
> filepath <- "C:/Multi_level_Illumina_feb2011/log2scale.normalized.txt" > dt <- read.delim(filepath, header = TRUE, as.is = TRUE, row.names = NULL)[,-1] > eSet <- dt[, grep("Signal", names(dt))] > seSet <- dt[, grep("STDERR", names(dt))] > nSet <- dt[, grep("NBEADS", names(dt))] > ProbeID <- dt[, 2] > iccResults <- iccFun(eSet, seSet, nSet, ProbeID = ProbeID, iccQuant = c(0, 0.25, 0.5, 0.75, 0.8, 1), diffIcc = TRUE, keepData = TRUE)Processing a LumiBatch object
> setwd("C:/Multi_level_Illumina_feb2011/RPackageFinal/beadstudiooutputData") # Read in the data using \code{lumiR} to obtain a LumiBatch object > x.lumi <- lumiR("AsuragenMAQC-probe-raw.txt") > lumi.N <- lumiN(x.lumi, "rsn") > eSet <- exprs(lumi.N) > seSet <- se.exprs(lumi.N) > nSet <- beadNum(lumi.N) > group <- c(1:dim(eSet)[2]) > ProbeID = fData(lumi.N)$ProbeID > iccResults <- iccFun(eSet, seSet, nSet, ProbeID = ProbeID, + iccQuant = c(0, 0.25, 0.5, 1), + diffIcc = TRUE, keepData = TRUE)Processing an ExpressionSetIllumina object
> exampleSummaryDataNorm <- + normaliseIllumina(channel(exampleSummaryData, "G"), + method = "quantile", transform = "none") > aaa <- + na.omit(data.frame(I(rownames(exprs(exampleSummaryDataNorm))), + exprs(exampleSummaryDataNorm))) > ProbeID <- aaa[, 1] > eSet <- na.omit(exprs(exampleSummaryDataNorm)) > stddev <- na.omit(se.exprs(exampleSummaryDataNorm)) > nSet <- na.omit(attributes(exampleSummaryDataNorm)$assayData$nObservations) > seSet <- stddev/sqrt(nSet) > iccResults <- iccFun(eSet, seSet, nSet, ProbeID = ProbeID, + iccQuant = c(0, 0.25, 0.5, 1))
The output of the iccFun function
In this subsection, we illustrate how the output of the
iccFun function can be observed. Note that we display the
results for exampleSummaryData, the output for the spike-in
data can be found in Forcheh, G. Verbeke,
A. Kasim, D. Lin, Z. Shkedy, W. Talloen, H. W. Göhlmann, and L. Clement
(2012).
> head(iccResults$betweenvar)
ProbeID fit1.tau2
1 ILMN_1802380 1.3154886475
2 ILMN_1893287 0.0202718744
3 ILMN_1736104 0.7883136626
4 ILMN_1792389 0.5374776179
5 ILMN_1854015 0.0000000000
6 ILMN_1904757 0.0004272419
> iccResults$withinvar[1:6, 1:6]
ProbeID sigma2.4613710017_B sigma2.4613710052_B sigma2.4613710054_B
ILMN_1802380 ILMN_1802380 0.08024396 0.1133679 0.07562057
ILMN_1893287 ILMN_1893287 0.15510050 0.1495736 0.31645854
ILMN_1736104 ILMN_1736104 0.22109680 0.2449570 0.16237022
ILMN_1792389 ILMN_1792389 0.16305881 0.2232660 0.25316536
ILMN_1854015 ILMN_1854015 0.31302729 0.1367953 0.29684239
ILMN_1904757 ILMN_1904757 0.11065525 0.2427457 0.35319329
sigma2.4616443079_B sigma2.4616443093_B
ILMN_1802380 0.1715118 0.1282700
ILMN_1893287 0.4629203 0.1956166
ILMN_1736104 0.2781976 0.2364219
ILMN_1792389 0.1187983 0.1560972
ILMN_1854015 0.2776505 0.4338320
ILMN_1904757 0.2621982 0.4215812
> head(iccResults$iccAll)
[,1] [,2] [,3] [,4] [,5] [,6]
[1,] 0.942507641 0.920658321 0.945640089 0.884659197 0.911155532 0.945550561
[2,] 0.115593318 0.119354803 0.060202088 0.041954058 0.093899753 0.076838800
[3,] 0.780964425 0.762930437 0.829206926 0.739151757 0.769284994 0.672257600
[4,] 0.767237213 0.706516107 0.679798133 0.818981163 0.774938161 0.794058983
[5,] 0.000000000 0.000000000 0.000000000 0.000000000 0.000000000 0.000000000
[6,] 0.003846168 0.001756947 0.001208193 0.001626811 0.001012401 0.003354805
[,7] [,8] [,9] [,10] [,11] [,12]
[1,] 0.813348960 0.2851197747 0.924933768 0.921820518 0.953747680 0.9672588084
[2,] 0.146496212 0.0086476707 0.079882094 0.046679979 0.083112294 0.0577318494
[3,] 0.862368982 0.4039671466 0.788168517 0.745235220 0.651296881 0.5468796285
[4,] 0.795226413 0.2446849630 0.818919242 0.889560842 0.859993713 0.7440351650
[5,] 0.000000000 0.0000000000 0.000000000 0.000000000 0.000000000 0.0000000000
[6,] 0.002726754 0.0005326228 0.001815729 0.005423024 0.002744552 0.0009235276
> head(iccResults$icc)
ProbeID q0 q0.25 q0.5 q1
1 ILMN_1802380 0.2851197747 0.904531448 0.923377143 0.967258808
2 ILMN_1893287 0.0086476707 0.054968882 0.078360447 0.146496212
3 ILMN_1736104 0.4039671466 0.667017420 0.754082829 0.862368982
4 ILMN_1792389 0.2446849630 0.734655400 0.784498572 0.889560842
5 ILMN_1854015 0.0000000000 0.000000000 0.000000000 0.000000000
6 ILMN_1904757 0.0005326228 0.001159245 0.001786338 0.005423024If desired, the iccResults$icc output from the
iccFun function can be used for filtering and assessing the
different filtering strategies.
# Obtaining the number of informative bead types at each of the specified ICC quantiles
> apply(iccResults$icc[, -1], 2, function(x, thres) sum(x >= thres), thres = 0.5)
q0 q0.25 q0.5 q1
4699 15784 17757 23419
# Obtaining the informative bead types using the minimum ICC
> filterDataNorm <- exampleSummaryDataNorm[subset(iccResults$icc,
+ iccResults$icc[, 2] >= 0.5)[, 1], ]
> dim(filterDataNorm)
Features Samples Channels
4699 12 1Assessing the filtering strategies
This is done using the emCDF function (Figure 1).
> emCDF(iccResults, iccQuant = c(0, 0.25, 0.5, 1))
Further, the within- and between-array variances at the minimum ICC
can be observed using the varianceplot function (Figure 2).
> varianceplot(iccResults, q = 1, delta = 0.8)
By specifying q = 1 and delta = 0.8, the
informative beads (displayed in red) are obtained using the minimum ICC
at a cutoff of 80%.
Analysis of differential expression
Once the data have been filtered, they can be used for downstream
analysis. Here, we assess differential expression using a filtered
data.frame. For the example 608 bead types were declared informative
based on the maximum ICC. We refer to the help file of the
MLM.beadarray function in the beadarrayFilter
package for an example on a ExpressionSetIllumina object.
> iccResults <- beadtypeFilter(dt, Quantile = 1, keepData = TRUE, delta = 0.5)
> dim(iccResults$informData)
> dat <- iccResults$informData
> eSet <- dat[, grep("Signal", names(dat))]
> seSet <- dat[, grep("STDERR", names(dat))]
> nSet <- dat[, grep("NBEADS", names(dat))]We define the group variable to compare concentrations 0.3 and 0.1 pM in the spike-in data. This is done by selecting the column numbers of the arrays corresponding to the concentrations of interest.
> group1 <- c(26, 32, 38, 44)
> group2 <- c(27, 33, 39, 45)
> fit1 <- MLM.beadarray(eSet, seSet, nSet, list(group1, group2),
+ var.equal = TRUE, max.iteration = 20, method = "ML")The output of the MLM.beadarray function can then be
used to test for equality of mean expression between the two
concentrations
> df <- length(group1) + length(group2) - 2
> fit1$pvalue <- 2 * (1-pt(abs(fit1$t.statistics), df))
> fit1$pvalAdjust <- p.adjust(fit1$pvalue, method = "fdr", n = length(fit1$pvalue))
> length(which(fit1$pvalAdjust < 0.05))
[1] 29i.e., 29 bead types were found to be differentially expressed between concentrations 0.3 and 0.1 pM. Note that 22 of these 29 bead types are true positives (spikes).
Discussion
The beadarrayFilter package can be used to filter Illumina
bead array data. The beadtypeFilter function can filter
normalized ExpressionSetIllumina objects, normalized LumiBatch objects
as well as normalized data.frames and returns the names of the
informative bead types. Optionally, the user can also obtain the
filtered data. This, however, does not return the required outputs to
assess different filtering strategies nor the variances using the
emCDF or the varianceplot functions,
respectively. The iccFun function can be used to customize
filtering strategies. It returns the required outputs for assessing
different filtering strategies and the between- and within-array
variances.
Acknowledgements
We acknowledge the support from IAP research network grant nr. P6/03 of the Belgian government (Belgian Science Policy), SymBioSys, the Katholieke Universiteit Leuven center of Excellence on Computational Systems Biology, (EF/05/007), and Bioframe of the institute for the Promotion of Innovation by Science and technology in Flanders (IWT: 060045/KUL-BIO-M$S-PLANT).
We are also grateful to (Kim and J. Lin
2011) for making the MLM.beadarray function
available.