MultiAssayExperiment classMultiAssayExperiment object: a rich example[GenomicRangesMultiAssayExperimentrownames and colnamessource("https://bioconductor.org/biocLite.R")
BiocInstaller::biocLite("MultiAssayExperiment")Loading the packages:
library(MultiAssayExperiment)
library(GenomicRanges)MultiAssayExperiment classHere is an overview of the class and its constructors and extractors:
empty <- MultiAssayExperiment()
empty## A MultiAssayExperiment object of 0 listed
## experiments with no user-defined names and respective classes.
## Containing an ExperimentList class object of length 0:
## Features:
## experiments() - obtain the ExperimentList instance
## colData() - the primary/phenotype DataFrame
## sampleMap() - the sample availability DataFrame
## `$`, `[`, `[[` - extract colData columns, subset, or experiment
## *Format() - convert into a long or wide DataFrame
## assays() - convert ExperimentList to a SimpleList of matricesslotNames(empty)## [1] "ExperimentList" "colData" "sampleMap" "metadata"
## [5] "drops"MultiAssayExperimentExperimentList: experimental dataThe ExperimentList slot and class is the container workhorse for the MultiAssayExperiment class. It contains all the experimental data. It inherits from class S4Vectors::SimpleList with one element/component per data type.
class(experiments(empty)) # ExperimentList## [1] "ExperimentList"
## attr(,"package")
## [1] "MultiAssayExperiment"The elements of the ExperimentList can contain ID-based and range-based data. Requirements for all classes in the ExperimentList are listed in the API.
The following base and Bioconductor classes are known to work as elements of the ExperimentList:
base::matrix: the base class, can be used for ID-based datasets such as gene expression summarized per-gene, microRNA, metabolomics, or microbiome data.
Biobase::ExpressionSet: A richer representation of ID-based datasets capable of storing additional assay-level metadata.
SummarizedExperiment::SummarizedExperiment: Also provides a rich representation of ID-based matrix-like datasets.
SummarizedExperiment::RangedSummarizedExperiment: For rectangular range-based datasets, one set of genomic ranges are assayed for multiple samples. It can be used for gene expression, methylation, or other data types that refer to genomic positions.
RaggedExperiment::RaggedExperiment: For range-based datasets, such as copy number and mutation data, the RaggedExperiment class can be used to represent measurements by genomic positions.
ExperimentList containerSee the API section for details on requirements for using other data classes. In general, data classes meeting minimum requirements, including support for square bracket [ subsetting and dimnames() will work by default.
The datasets contained in elements of the ExperimentList can have:
The column names correspond to samples, and are used to match assay data to specimen metadata stored in colData.
The row names can correspond to a variety of features in the data including but not limited to gene names, probe IDs, proteins, and named ranges. Note that the existence of “row” names does not mean the data must be rectangular or matrix-like.
Classes contained in the ExperimentList must support the following list of methods:
[: single square bracket subsetting, with a single comma. It is assumed that values before the comma subset rows, and values after the comma subset columns.dimnames() : corresponding to features (such as genes, proteins, etc.) and experimental samplesdim(): returns a vector of the number of rows and number of columnscolData: primary dataThe MultiAssayExperiment keeps one set of “primary” metadata that describes the ‘biological unit’ which can refer to specimens, experimental subjects, patients, etc. In this vignette, we will refer to each experimental subject as a patient.
colData slot requirementsThe colData dataset should be of class DataFrame but can accept a data.frame class object that will be coerced.
In order to relate metadata of the biological unit, the row names of the colData dataset must contain patient identifiers.
patient.data <- data.frame(sex=c("M", "F", "M", "F"),
age=38:41,
row.names=c("Jack", "Jill", "Bob", "Barbara"))
patient.data## sex age
## Jack M 38
## Jill F 39
## Bob M 40
## Barbara F 41DataFrameFor many typical purposes the DataFrame and data.frame behave equivalently; but the Dataframe is more flexible as it allows any vector-like data type to be stored in its columns. The flexibility of the DataFrame permits, for example, storing multiple dose-response values for a single cell line, even if the number of doses and responses is not consistent across all cell lines. Doses could be stored in one column of colData as a SimpleList, and responses in another column, also as a SimpleList. Or, dose-response values could be stored in a single column of colData as a two-column matrix for each cell line.
sampleMap: relating colData to multiple assaysThe sampleMap is a DataFrame that relates the “primary” data (colData) to the experimental assays:
class(sampleMap(empty)) # DataFrame## [1] "DataFrame"
## attr(,"package")
## [1] "S4Vectors"The sampleMap provides an unambiguous map from every experimental observation to one and only one row in colData. It is, however, permissible for a row of colData to be associated with multiple experimental observations or no observations at all. In other words, there is a “many-to-one” mapping from experimental observations to rows of colData, and a “one-to-any-number” mapping from rows of colData to experimental observations.
sampleMap structureThe sampleMap has three columns, with the following column names:
assay provides the names of the different experiments / assays performed. These are user-defined, with the only requirement that the names of the ExperimentList, where the experimental assays are stored, must be contained in this column.
primary provides the “primary” sample names. All values in this column must also be present in the rownames of colData(MultiAssayExperiment). In this example, allowable values in this column are “Jack”, “Jill”, “Barbara”, and “Bob”.
colname provides the sample names used by experimental datasets, which in practice are often different than the primary sample names. For each assay, all column names must be found in this column. Otherwise, those assays would be orphaned: it would be impossible to match them up to samples in the overall experiment. As mentioned above, duplicated values are allowed, to represent replicates with the same overall experiment-level annotation.
This design is motivated by the following situations:
sampleMap isn’t providedIf each assay uses the same colnames (i.e., if the same sample identifiers are used for each experiment), a simple list of these datasets is sufficient for the MultiAssayExperiment constructor function. It is not necessary for them to have the same rownames or colnames:
exprss1 <- matrix(rnorm(16), ncol = 4,
dimnames = list(sprintf("ENST00000%i", sample(288754:290000, 4)),
c("Jack", "Jill", "Bob", "Bobby")))
exprss2 <- matrix(rnorm(12), ncol = 3,
dimnames = list(sprintf("ENST00000%i", sample(288754:290000, 4)),
c("Jack", "Jane", "Bob")))
doubleExp <- list("methyl 2k" = exprss1, "methyl 3k" = exprss2)
simpleMultiAssay <- MultiAssayExperiment(experiments=doubleExp)
simpleMultiAssay## A MultiAssayExperiment object of 2 listed
## experiments with user-defined names and respective classes.
## Containing an ExperimentList class object of length 2:
## [1] methyl 2k: matrix with 4 rows and 4 columns
## [2] methyl 3k: matrix with 4 rows and 3 columns
## Features:
## experiments() - obtain the ExperimentList instance
## colData() - the primary/phenotype DataFrame
## sampleMap() - the sample availability DataFrame
## `$`, `[`, `[[` - extract colData columns, subset, or experiment
## *Format() - convert into a long or wide DataFrame
## assays() - convert ExperimentList to a SimpleList of matricesIn the above example, the user did not provide the colData argument so the constructor function filled it with an empty DataFrame:
colData(simpleMultiAssay)## DataFrame with 5 rows and 0 columnsBut the colData can be provided. Here, note that any assay sample (column) that cannot be mapped to a corresponding row in the provided colData gets dropped. This is part of ensuring internal validity of the MultiAssayExperiment.
simpleMultiAssay2 <- MultiAssayExperiment(experiments=doubleExp,
colData=patient.data)## Warning in .generateMap(colData, experiments): Data from rows:
## NA - Bobby
## NA - Jane
## dropped due to missing phenotype data## harmonizing input:
## removing 1 colData rownames not in sampleMap 'primary'simpleMultiAssay2## A MultiAssayExperiment object of 2 listed
## experiments with user-defined names and respective classes.
## Containing an ExperimentList class object of length 2:
## [1] methyl 2k: matrix with 4 rows and 3 columns
## [2] methyl 3k: matrix with 4 rows and 2 columns
## Features:
## experiments() - obtain the ExperimentList instance
## colData() - the primary/phenotype DataFrame
## sampleMap() - the sample availability DataFrame
## `$`, `[`, `[[` - extract colData columns, subset, or experiment
## *Format() - convert into a long or wide DataFrame
## assays() - convert ExperimentList to a SimpleList of matricescolData(simpleMultiAssay2)## DataFrame with 3 rows and 2 columns
## sex age
## <factor> <integer>
## Jack M 38
## Jill F 39
## Bob M 40Metadata can be added at different levels of the MultiAssayExperiment.
Can be of ANY class, for storing study-wide metadata, such as citation information. For an empty MultiAssayExperiment object, it is NULL.
class(metadata(empty)) # NULL (class "ANY")## [1] "NULL"At the ExperimentList level, the metadata function would allow the user to enter metadata as a list.
metadata(experiments(empty))## list()At the individual assay level, certain classes may support metadata, for example, metadata and mcols for a SummarizedExperiment. It is recommended to use metadata at the ExperimentList level.
MultiAssayExperiment object: a rich exampleIn this section we demonstrate all core supported data classes, using different sample ID conventions for each assay, with primary colData. The some supported classes such as, matrix, ExpressionSet, SummarizedExperiment, and RangedSummarizedExperiment.
We have three matrix-like datasets. First, let’s represent expression data as an ExpressionSet:
library(Biobase)
(arraydat <- matrix(seq(101, 108), ncol=4,
dimnames=list(c("ENST00000294241", "ENST00000355076"),
c("array1", "array2", "array3", "array4"))))## array1 array2 array3 array4
## ENST00000294241 101 103 105 107
## ENST00000355076 102 104 106 108arraypdat <- as(data.frame(slope53=rnorm(4),
row.names=c("array1", "array2", "array3",
"array4")), "AnnotatedDataFrame")
exprdat <- ExpressionSet(assayData=arraydat, phenoData=arraypdat)
exprdat## ExpressionSet (storageMode: lockedEnvironment)
## assayData: 2 features, 4 samples
## element names: exprs
## protocolData: none
## phenoData
## sampleNames: array1 array2 array3 array4
## varLabels: slope53
## varMetadata: labelDescription
## featureData: none
## experimentData: use 'experimentData(object)'
## Annotation:The following map matches colData sample names to exprdata sample names. Note that row orders aren’t initially matched up, and this is OK.
(exprmap <- data.frame(primary=rownames(patient.data)[c(1, 2, 4, 3)],
assay=c("array1", "array2", "array3", "array4"),
stringsAsFactors = FALSE))## primary assay
## 1 Jack array1
## 2 Jill array2
## 3 Barbara array3
## 4 Bob array4Now methylation data, which we will represent as a matrix. It uses gene identifiers also, but measures a partially overlapping set of genes. Now, let’s store this as a simple matrix which can contains a replicate for one of the patients.
(methyldat <-
matrix(1:10, ncol=5,
dimnames=list(c("ENST00000355076", "ENST00000383706"),
c("methyl1", "methyl2", "methyl3",
"methyl4", "methyl5"))))## methyl1 methyl2 methyl3 methyl4 methyl5
## ENST00000355076 1 3 5 7 9
## ENST00000383706 2 4 6 8 10The following map matches colData sample names to methyldat sample names.
(methylmap <- data.frame(primary = c("Jack", "Jack", "Jill", "Barbara", "Bob"),
assay = c("methyl1", "methyl2", "methyl3", "methyl4", "methyl5"),
stringsAsFactors = FALSE))## primary assay
## 1 Jack methyl1
## 2 Jack methyl2
## 3 Jill methyl3
## 4 Barbara methyl4
## 5 Bob methyl5Now we have a microRNA platform, which has no common identifiers with the other datasets, and which we also represent as a matrix. It is also missing data for “Jill”. We will use the same sample naming convention as we did for arrays.
(microdat <- matrix(201:212, ncol=3,
dimnames=list(c("hsa-miR-21", "hsa-miR-191",
"hsa-miR-148a", "hsa-miR148b"),
c("micro1", "micro2", "micro3"))))## micro1 micro2 micro3
## hsa-miR-21 201 205 209
## hsa-miR-191 202 206 210
## hsa-miR-148a 203 207 211
## hsa-miR148b 204 208 212And the following map matches colData sample names to microdat sample names.
(micromap <- data.frame(primary = c("Jack", "Barbara", "Bob"),
assay = c("micro1", "micro2", "micro3"),
stringsAsFactors = FALSE))## primary assay
## 1 Jack micro1
## 2 Barbara micro2
## 3 Bob micro3Finally, we create a dataset of class RangedSummarizedExperiment:
library(SummarizedExperiment)## Loading required package: DelayedArray## Loading required package: matrixStats##
## Attaching package: 'matrixStats'## The following objects are masked from 'package:Biobase':
##
## anyMissing, rowMedians##
## Attaching package: 'DelayedArray'## The following objects are masked from 'package:matrixStats':
##
## colMaxs, colMins, colRanges, rowMaxs, rowMins, rowRanges## The following object is masked from 'package:base':
##
## applynrows <- 5; ncols <- 4
counts <- matrix(runif(nrows * ncols, 1, 1e4), nrows)
rowRanges <- GRanges(rep(c("chr1", "chr2"), c(2, nrows - 2)),
IRanges(floor(runif(nrows, 1e5, 1e6)), width=100),
strand=sample(c("+", "-"), nrows, TRUE),
feature_id=sprintf("ID\\%03d", 1:nrows))
names(rowRanges) <- letters[1:5]
colData <- DataFrame(Treatment=rep(c("ChIP", "Input"), 2),
row.names= c("mysnparray1", "mysnparray2",
"mysnparray3", "mysnparray4"))
rse <- SummarizedExperiment(assays=SimpleList(counts=counts),
rowRanges=rowRanges, colData=colData)And we map the colData samples to the RangedSummarizedExperiment:
(rangemap <-
data.frame(primary = c("Jack", "Jill", "Bob", "Barbara"),
assay = c("mysnparray1", "mysnparray2", "mysnparray3",
"mysnparray4"), stringsAsFactors = FALSE))## primary assay
## 1 Jack mysnparray1
## 2 Jill mysnparray2
## 3 Bob mysnparray3
## 4 Barbara mysnparray4sampleMap creationThe MultiAssayExperiment constructor function can create the sampleMap automatically if a single naming convention is used, but in this example it cannot because we used platform-specific sample identifiers (e.g. mysnparray1, etc). So we must provide an ID map that matches the samples of each experiment back to the colData, as a three-column data.frame or DataFrame with three columns named “assay”, primary“, and”colname“. Here we start with a list:
listmap <- list(exprmap, methylmap, micromap, rangemap)
names(listmap) <- c("Affy", "Methyl 450k", "Mirna", "CNV gistic")
listmap## $Affy
## primary assay
## 1 Jack array1
## 2 Jill array2
## 3 Barbara array3
## 4 Bob array4
##
## $`Methyl 450k`
## primary assay
## 1 Jack methyl1
## 2 Jack methyl2
## 3 Jill methyl3
## 4 Barbara methyl4
## 5 Bob methyl5
##
## $Mirna
## primary assay
## 1 Jack micro1
## 2 Barbara micro2
## 3 Bob micro3
##
## $`CNV gistic`
## primary assay
## 1 Jack mysnparray1
## 2 Jill mysnparray2
## 3 Bob mysnparray3
## 4 Barbara mysnparray4and use the convenience function listToMap to convert the list of data.frame objects to a valid object for the sampleMap:
dfmap <- listToMap(listmap)
dfmap## DataFrame with 16 rows and 3 columns
## assay primary colname
## <factor> <character> <character>
## 1 Affy Jack array1
## 2 Affy Jill array2
## 3 Affy Barbara array3
## 4 Affy Bob array4
## 5 Methyl 450k Jack methyl1
## ... ... ... ...
## 12 Mirna Bob micro3
## 13 CNV gistic Jack mysnparray1
## 14 CNV gistic Jill mysnparray2
## 15 CNV gistic Bob mysnparray3
## 16 CNV gistic Barbara mysnparray4Note, dfmap can be reverted to a list with another provided function:
mapToList(dfmap, "assay")list()Create an named list of experiments for the MultiAssayExperiment function. All of these names must be found within in the third column of dfmap:
objlist <- list("Affy" = exprdat, "Methyl 450k" = methyldat,
"Mirna" = microdat, "CNV gistic" = rse)MultiAssayExperiment class objectWe recommend using the MultiAssayExperiment constructor function:
myMultiAssay <- MultiAssayExperiment(objlist, patient.data, dfmap)
myMultiAssay## A MultiAssayExperiment object of 4 listed
## experiments with user-defined names and respective classes.
## Containing an ExperimentList class object of length 4:
## [1] Affy: ExpressionSet with 2 rows and 4 columns
## [2] Methyl 450k: matrix with 2 rows and 5 columns
## [3] Mirna: matrix with 4 rows and 3 columns
## [4] CNV gistic: RangedSummarizedExperiment with 5 rows and 4 columns
## Features:
## experiments() - obtain the ExperimentList instance
## colData() - the primary/phenotype DataFrame
## sampleMap() - the sample availability DataFrame
## `$`, `[`, `[[` - extract colData columns, subset, or experiment
## *Format() - convert into a long or wide DataFrame
## assays() - convert ExperimentList to a SimpleList of matricesThe following extractor functions can be used to get extract data from the object:
experiments(myMultiAssay)## ExperimentList class object of length 4:
## [1] Affy: ExpressionSet with 2 rows and 4 columns
## [2] Methyl 450k: matrix with 2 rows and 5 columns
## [3] Mirna: matrix with 4 rows and 3 columns
## [4] CNV gistic: RangedSummarizedExperiment with 5 rows and 4 columnscolData(myMultiAssay)## DataFrame with 4 rows and 2 columns
## sex age
## <factor> <integer>
## Jack M 38
## Jill F 39
## Bob M 40
## Barbara F 41sampleMap(myMultiAssay)## DataFrame with 16 rows and 3 columns
## assay primary colname
## <factor> <character> <character>
## 1 Affy Jack array1
## 2 Affy Jill array2
## 3 Affy Barbara array3
## 4 Affy Bob array4
## 5 Methyl 450k Jack methyl1
## ... ... ... ...
## 12 Mirna Bob micro3
## 13 CNV gistic Jack mysnparray1
## 14 CNV gistic Jill mysnparray2
## 15 CNV gistic Bob mysnparray3
## 16 CNV gistic Barbara mysnparray4metadata(myMultiAssay)## NULLNote that the ExperimentList class extends the SimpleList class to add some validity checks specific to MultiAssayExperiment. It can be used like a list.
MultiAssayExperiment objectThe prepMultiAssay function helps diagnose common problems when creating a MultiAssayExperiment object. It provides error messages and/or warnings in instances where names (either colnames or ExperimentList element names) are inconsistent with those found in the sampleMap. Input arguments are the same as those in the MultiAssayExperiment (i.e., ExperimentList, colData, sampleMap). The resulting output of the prepMultiAssay function is a list of inputs including a “metadata$drops” element for names that were not able to be matched.
Instances where ExperimentList is created without names will prompt an error from prepMultiAssay. Named ExperimentList elements are essential for checks in MultiAssayExperiment.
objlist3 <- objlist
(names(objlist3) <- NULL)## NULLtry(prepMultiAssay(objlist3, patient.data, dfmap)$experiments)Non-matching names may also be present in the ExperimentList elements and the “assay” column of the sampleMap. If names only differ by case and are identical and unique, names will be standardized to lower case and replaced.
names(objlist3) <- toupper(names(objlist))
names(objlist3)## [1] "AFFY" "METHYL 450K" "MIRNA" "CNV GISTIC"unique(dfmap[, "assay"])## [1] Affy Methyl 450k Mirna CNV gistic
## Levels: Affy Methyl 450k Mirna CNV gisticprepMultiAssay(objlist3, patient.data, dfmap)$experiments##
## Names in the ExperimentList do not match sampleMap assay
## standardizing will be attempted...## - names set to lowercase## ExperimentList class object of length 4:
## [1] affy: ExpressionSet with 2 rows and 4 columns
## [2] methyl 450k: matrix with 2 rows and 5 columns
## [3] mirna: matrix with 4 rows and 3 columns
## [4] cnv gistic: RangedSummarizedExperiment with 5 rows and 4 columnsWhen colnames in the ExperimentList cannot be matched back to the primary data (colData), these will be dropped and added to the drops element.
exampleMap <- sampleMap(simpleMultiAssay2)
sapply(doubleExp, colnames)## $`methyl 2k`
## [1] "Jack" "Jill" "Bob" "Bobby"
##
## $`methyl 3k`
## [1] "Jack" "Jane" "Bob"exampleMap## DataFrame with 5 rows and 3 columns
## assay primary colname
## <factor> <character> <character>
## 1 methyl 2k Jack Jack
## 2 methyl 2k Jill Jill
## 3 methyl 2k Bob Bob
## 4 methyl 3k Jack Jack
## 5 methyl 3k Bob BobprepMultiAssay(doubleExp, patient.data, exampleMap)$metadata$drops##
## Not all colnames in the ExperimentList are found in the
## sampleMap, dropping samples from ExperimentList...## $`methyl 2k`
## [1] "Bobby"
##
## $`methyl 3k`
## [1] "Jane"## $`columns.methyl 2k`
## [1] "Bobby"
##
## $`columns.methyl 3k`
## [1] "Jane"A similar operation is performed for checking “primary” sampleMap names and colData rownames. In this example, we add a row corresponding to “Joe” that does not have a match in the experimental data.
exMap <- rbind(dfmap,
DataFrame(assay = "New methyl",
primary = "Joe",
colname = "Joe"))
invisible(prepMultiAssay(objlist, patient.data, exMap))## Warning in prepMultiAssay(objlist, patient.data, exMap):
## Lengths of names in the ExperimentList and sampleMap
## are not equal##
## Not all names in the primary column of the sampleMap
## could be matched to the colData rownames; see $drops## DataFrame with 1 row and 3 columns
## assay primary colname
## <factor> <character> <character>
## 1 New methyl Joe JoeTo create a MultiAssayExperiment from the results of the prepMultiAssay function, take each corresponding element from the resulting list and enter them as arguments to the MultiAssayExperiment constructor function.
prepped <- prepMultiAssay(objlist, patient.data, exMap)## Warning in prepMultiAssay(objlist, patient.data, exMap):
## Lengths of names in the ExperimentList and sampleMap
## are not equal##
## Not all names in the primary column of the sampleMap
## could be matched to the colData rownames; see $drops## DataFrame with 1 row and 3 columns
## assay primary colname
## <factor> <character> <character>
## 1 New methyl Joe JoepreppedMulti <- MultiAssayExperiment(prepped$experiments, prepped$colData,
prepped$sampleMap, prepped$metadata)
preppedMulti## A MultiAssayExperiment object of 4 listed
## experiments with user-defined names and respective classes.
## Containing an ExperimentList class object of length 4:
## [1] Affy: ExpressionSet with 2 rows and 4 columns
## [2] Methyl 450k: matrix with 2 rows and 5 columns
## [3] Mirna: matrix with 4 rows and 3 columns
## [4] CNV gistic: RangedSummarizedExperiment with 5 rows and 4 columns
## Features:
## experiments() - obtain the ExperimentList instance
## colData() - the primary/phenotype DataFrame
## sampleMap() - the sample availability DataFrame
## `$`, `[`, `[[` - extract colData columns, subset, or experiment
## *Format() - convert into a long or wide DataFrame
## assays() - convert ExperimentList to a SimpleList of matricesAlternatively, use the do.call function to easily create a MultiAssayExperiment from the output of prepMultiAssay function:
do.call(MultiAssayExperiment, prepped)## A MultiAssayExperiment object of 4 listed
## experiments with user-defined names and respective classes.
## Containing an ExperimentList class object of length 4:
## [1] Affy: ExpressionSet with 2 rows and 4 columns
## [2] Methyl 450k: matrix with 2 rows and 5 columns
## [3] Mirna: matrix with 4 rows and 3 columns
## [4] CNV gistic: RangedSummarizedExperiment with 5 rows and 4 columns
## Features:
## experiments() - obtain the ExperimentList instance
## colData() - the primary/phenotype DataFrame
## sampleMap() - the sample availability DataFrame
## `$`, `[`, `[[` - extract colData columns, subset, or experiment
## *Format() - convert into a long or wide DataFrame
## assays() - convert ExperimentList to a SimpleList of matricesBioconductor classes from raw dataRecent updates to the GenomicRanges and SummarizedExperiment packages allow the user to create standard Bioconductor classes from raw data. Raw data read in as either data.frame or DataFrame can be converted to GRangesList or SummarizedExperiment classes depending on the type of data.
The function to create a GRangesList from a data.frame, called makeGRangesListFromDataFrame can be found in the GenomicRanges package. makeSummarizedExperimentFromDataFrame is available in the SummarizedExperiment package. It is also possible to create a RangedSummarizedExperiment class object from raw data when ranged data is available.
A simple example can be obtained from the function documentation in GenomicRanges:
library(GenomicRanges)
grlls <- list(chr = rep("chr1", nrows), start = seq(11, 15),
end = seq(12, 16), strand = c("+", "-", "+", "*", "*"),
score = seq(1, 5), specimen = c("a", "a", "b", "b", "c"),
gene_symbols = paste0("GENE", letters[seq_len(nrows)]))
grldf <- as.data.frame(grlls, stringsAsFactors = FALSE)
GRL <- makeGRangesListFromDataFrame(grldf, split.field = "specimen",
names.field = "gene_symbols")This can then be converted to a RaggedExperiment object for a rectangular representation that will conform more easily to the MultiAssayExperiment API requirements.
library(RaggedExperiment)## in method for 'coerce' with signature '"RangedRaggedAssay","RaggedExperiment"': no definition for class "RaggedExperiment"RaggedExperiment(GRL)## class: RaggedExperiment
## dim: 5 3
## assays(0):
## rownames(5): GENEa GENEb GENEc GENEd GENEe
## colnames(3): a b c
## colData names(0):Note. See the RaggedExperiment vignette for more details.
In the SummarizedExperiment package:
library(SummarizedExperiment)
sels <- list(chr = rep("chr2", nrows), start = seq(11, 15),
end = seq(12, 16), strand = c("+", "-", "+", "*", "*"),
expr0 = seq(3, 7), expr1 = seq(8, 12), expr2 = seq(12, 16))
sedf <- as.data.frame(sels, row.names = paste0("GENE",
letters[rev(seq_len(nrows))]),
stringsAsFactors = FALSE)
sedf## chr start end strand expr0 expr1 expr2
## GENEe chr2 11 12 + 3 8 12
## GENEd chr2 12 13 - 4 9 13
## GENEc chr2 13 14 + 5 10 14
## GENEb chr2 14 15 * 6 11 15
## GENEa chr2 15 16 * 7 12 16makeSummarizedExperimentFromDataFrame(sedf)## class: RangedSummarizedExperiment
## dim: 5 3
## metadata(0):
## assays(1): ''
## rownames(5): GENEe GENEd GENEc GENEb GENEa
## rowData names(0):
## colnames(3): expr0 expr1 expr2
## colData names(0):MultiAssayExperiment allows subsetting by rows, columns, and assays, rownames, and colnames, across all experiments simultaneously while guaranteeing continued matching of samples.
Subsetting can be done most compactly by the square bracket method, or more verbosely and potentially more flexibly by the subsetBy*() methods.
[The three positions within the bracket operator indicate rows, columns, and assays, respectively (pseudocode):
myMultiAssay[rows, columns, assays]For example, to select the gene “ENST00000355076”:
myMultiAssay["ENST00000355076", , ]## A MultiAssayExperiment object of 2 listed
## experiments with user-defined names and respective classes.
## Containing an ExperimentList class object of length 2:
## [1] Affy: ExpressionSet with 1 rows and 4 columns
## [2] Methyl 450k: matrix with 1 rows and 5 columns
## Features:
## experiments() - obtain the ExperimentList instance
## colData() - the primary/phenotype DataFrame
## sampleMap() - the sample availability DataFrame
## `$`, `[`, `[[` - extract colData columns, subset, or experiment
## *Format() - convert into a long or wide DataFrame
## assays() - convert ExperimentList to a SimpleList of matricesThe above operation works across all types of assays, whether ID-based (e.g. matrix, ExpressionSet, SummarizedExperiment) or range-based (e.g. RangedSummarizedExperiment). Note that when using the bracket method [, the drop argument is TRUE by default.
You can subset by rows, columns, and assays in a single bracket operation, and they will be performed in that order (rows, then columns, then assays). The following selects the ENST00000355076 gene across all samples, then the first two samples of each assay, and finally the Affy and Methyl 450k assays:
myMultiAssay["ENST00000355076", 1:2, c("Affy", "Methyl 450k")]## harmonizing input:
## removing 4 sampleMap rows with 'colname' not in colnames of experiments
## removing 2 colData rownames not in sampleMap 'primary'## A MultiAssayExperiment object of 2 listed
## experiments with user-defined names and respective classes.
## Containing an ExperimentList class object of length 2:
## [1] Affy: ExpressionSet with 1 rows and 2 columns
## [2] Methyl 450k: matrix with 1 rows and 3 columns
## Features:
## experiments() - obtain the ExperimentList instance
## colData() - the primary/phenotype DataFrame
## sampleMap() - the sample availability DataFrame
## `$`, `[`, `[[` - extract colData columns, subset, or experiment
## *Format() - convert into a long or wide DataFrame
## assays() - convert ExperimentList to a SimpleList of matricesBy columns - character, integer, and logical are all allowed, for example:
myMultiAssay[, "Jack", ]## harmonizing input:
## removing 11 sampleMap rows with 'colname' not in colnames of experiments
## removing 3 colData rownames not in sampleMap 'primary'## A MultiAssayExperiment object of 4 listed
## experiments with user-defined names and respective classes.
## Containing an ExperimentList class object of length 4:
## [1] Affy: ExpressionSet with 2 rows and 1 columns
## [2] Methyl 450k: matrix with 2 rows and 2 columns
## [3] Mirna: matrix with 4 rows and 1 columns
## [4] CNV gistic: RangedSummarizedExperiment with 5 rows and 1 columns
## Features:
## experiments() - obtain the ExperimentList instance
## colData() - the primary/phenotype DataFrame
## sampleMap() - the sample availability DataFrame
## `$`, `[`, `[[` - extract colData columns, subset, or experiment
## *Format() - convert into a long or wide DataFrame
## assays() - convert ExperimentList to a SimpleList of matricesmyMultiAssay[, 1, ]## harmonizing input:
## removing 11 sampleMap rows with 'colname' not in colnames of experiments
## removing 3 colData rownames not in sampleMap 'primary'## A MultiAssayExperiment object of 4 listed
## experiments with user-defined names and respective classes.
## Containing an ExperimentList class object of length 4:
## [1] Affy: ExpressionSet with 2 rows and 1 columns
## [2] Methyl 450k: matrix with 2 rows and 2 columns
## [3] Mirna: matrix with 4 rows and 1 columns
## [4] CNV gistic: RangedSummarizedExperiment with 5 rows and 1 columns
## Features:
## experiments() - obtain the ExperimentList instance
## colData() - the primary/phenotype DataFrame
## sampleMap() - the sample availability DataFrame
## `$`, `[`, `[[` - extract colData columns, subset, or experiment
## *Format() - convert into a long or wide DataFrame
## assays() - convert ExperimentList to a SimpleList of matricesmyMultiAssay[, c(TRUE, FALSE, FALSE, FALSE), ]## harmonizing input:
## removing 11 sampleMap rows with 'colname' not in colnames of experiments
## removing 3 colData rownames not in sampleMap 'primary'## A MultiAssayExperiment object of 4 listed
## experiments with user-defined names and respective classes.
## Containing an ExperimentList class object of length 4:
## [1] Affy: ExpressionSet with 2 rows and 1 columns
## [2] Methyl 450k: matrix with 2 rows and 2 columns
## [3] Mirna: matrix with 4 rows and 1 columns
## [4] CNV gistic: RangedSummarizedExperiment with 5 rows and 1 columns
## Features:
## experiments() - obtain the ExperimentList instance
## colData() - the primary/phenotype DataFrame
## sampleMap() - the sample availability DataFrame
## `$`, `[`, `[[` - extract colData columns, subset, or experiment
## *Format() - convert into a long or wide DataFrame
## assays() - convert ExperimentList to a SimpleList of matricesBy assay - character, integer, and logical are allowed:
myMultiAssay[, , "Mirna"]## harmonizing input:
## removing 1 colData rownames not in sampleMap 'primary'## A MultiAssayExperiment object of 1 listed
## experiment with a user-defined name and respective class.
## Containing an ExperimentList class object of length 1:
## [1] Mirna: matrix with 4 rows and 3 columns
## Features:
## experiments() - obtain the ExperimentList instance
## colData() - the primary/phenotype DataFrame
## sampleMap() - the sample availability DataFrame
## `$`, `[`, `[[` - extract colData columns, subset, or experiment
## *Format() - convert into a long or wide DataFrame
## assays() - convert ExperimentList to a SimpleList of matricesmyMultiAssay[, , 3]## harmonizing input:
## removing 1 colData rownames not in sampleMap 'primary'## A MultiAssayExperiment object of 1 listed
## experiment with a user-defined name and respective class.
## Containing an ExperimentList class object of length 1:
## [1] Mirna: matrix with 4 rows and 3 columns
## Features:
## experiments() - obtain the ExperimentList instance
## colData() - the primary/phenotype DataFrame
## sampleMap() - the sample availability DataFrame
## `$`, `[`, `[[` - extract colData columns, subset, or experiment
## *Format() - convert into a long or wide DataFrame
## assays() - convert ExperimentList to a SimpleList of matricesmyMultiAssay[, , c(FALSE, FALSE, TRUE, FALSE, FALSE)]## harmonizing input:
## removing 1 colData rownames not in sampleMap 'primary'## A MultiAssayExperiment object of 1 listed
## experiment with a user-defined name and respective class.
## Containing an ExperimentList class object of length 1:
## [1] Mirna: matrix with 4 rows and 3 columns
## Features:
## experiments() - obtain the ExperimentList instance
## colData() - the primary/phenotype DataFrame
## sampleMap() - the sample availability DataFrame
## `$`, `[`, `[[` - extract colData columns, subset, or experiment
## *Format() - convert into a long or wide DataFrame
## assays() - convert ExperimentList to a SimpleList of matricesSpecify drop=FALSE to keep assays with zero rows or zero columns, e.g.:
myMultiAssay["ENST00000355076", , , drop=FALSE]## A MultiAssayExperiment object of 4 listed
## experiments with user-defined names and respective classes.
## Containing an ExperimentList class object of length 4:
## [1] Affy: ExpressionSet with 1 rows and 4 columns
## [2] Methyl 450k: matrix with 1 rows and 5 columns
## [3] Mirna: matrix with 0 rows and 3 columns
## [4] CNV gistic: RangedSummarizedExperiment with 0 rows and 4 columns
## Features:
## experiments() - obtain the ExperimentList instance
## colData() - the primary/phenotype DataFrame
## sampleMap() - the sample availability DataFrame
## `$`, `[`, `[[` - extract colData columns, subset, or experiment
## *Format() - convert into a long or wide DataFrame
## assays() - convert ExperimentList to a SimpleList of matricesUsing the default drop=TRUE, assays with no rows or no columns are removed:
myMultiAssay["ENST00000355076", , , drop=TRUE]## A MultiAssayExperiment object of 2 listed
## experiments with user-defined names and respective classes.
## Containing an ExperimentList class object of length 2:
## [1] Affy: ExpressionSet with 1 rows and 4 columns
## [2] Methyl 450k: matrix with 1 rows and 5 columns
## Features:
## experiments() - obtain the ExperimentList instance
## colData() - the primary/phenotype DataFrame
## sampleMap() - the sample availability DataFrame
## `$`, `[`, `[[` - extract colData columns, subset, or experiment
## *Format() - convert into a long or wide DataFrame
## assays() - convert ExperimentList to a SimpleList of matricesExperimental samples are stored in the rows of colData but the columns of elements of ExperimentList, so when we refer to subsetting by columns, we are referring to columns of the experimental assays. Subsetting by samples / columns will be more obvious after recalling the colData:
colData(myMultiAssay)## DataFrame with 4 rows and 2 columns
## sex age
## <factor> <integer>
## Jack M 38
## Jill F 39
## Bob M 40
## Barbara F 41Subsetting by samples identifies the selected samples in rows of the colData DataFrame, then selects all columns of the ExperimentList corresponding to these rows. Here we use an integer to keep the first two rows of colData, and all experimental assays associated to those two primary samples:
myMultiAssay[, 1:2]## harmonizing input:
## removing 8 sampleMap rows with 'colname' not in colnames of experiments
## removing 2 colData rownames not in sampleMap 'primary'## A MultiAssayExperiment object of 4 listed
## experiments with user-defined names and respective classes.
## Containing an ExperimentList class object of length 4:
## [1] Affy: ExpressionSet with 2 rows and 2 columns
## [2] Methyl 450k: matrix with 2 rows and 3 columns
## [3] Mirna: matrix with 4 rows and 1 columns
## [4] CNV gistic: RangedSummarizedExperiment with 5 rows and 2 columns
## Features:
## experiments() - obtain the ExperimentList instance
## colData() - the primary/phenotype DataFrame
## sampleMap() - the sample availability DataFrame
## `$`, `[`, `[[` - extract colData columns, subset, or experiment
## *Format() - convert into a long or wide DataFrame
## assays() - convert ExperimentList to a SimpleList of matricesNote that the above operation keeps different numbers of columns / samples from each assay, reflecting the reality that some samples may not have been assayed in all experiments, and may have replicates in some.
Columns can be subset using a logical vector. Here the dollar sign operator ($) accesses one of the columns in colData.
malesMultiAssay <- myMultiAssay[, myMultiAssay$sex == "M"]## harmonizing input:
## removing 7 sampleMap rows with 'colname' not in colnames of experiments
## removing 2 colData rownames not in sampleMap 'primary'colData(malesMultiAssay)## DataFrame with 2 rows and 2 columns
## sex age
## <factor> <integer>
## Jack M 38
## Bob M 40Finally, for special use cases you can exert detailed control of row or column subsetting, by using a list or CharacterList to subset. The following creates a CharacterList of the column names of each assay:
allsamples <- colnames(myMultiAssay)
allsamples## CharacterList of length 4
## [["Affy"]] array1 array2 array3 array4
## [["Methyl 450k"]] methyl1 methyl2 methyl3 methyl4 methyl5
## [["Mirna"]] micro1 micro2 micro3
## [["CNV gistic"]] mysnparray1 mysnparray2 mysnparray3 mysnparray4Now let’s get rid of three Methyl 450k arrays, those in positions 3, 4, and 5:
allsamples[["Methyl 450k"]] <- allsamples[["Methyl 450k"]][-3:-5]
myMultiAssay[, as.list(allsamples), ]## harmonizing input:
## removing 3 sampleMap rows with 'colname' not in colnames of experiments## A MultiAssayExperiment object of 4 listed
## experiments with user-defined names and respective classes.
## Containing an ExperimentList class object of length 4:
## [1] Affy: ExpressionSet with 2 rows and 4 columns
## [2] Methyl 450k: matrix with 2 rows and 2 columns
## [3] Mirna: matrix with 4 rows and 3 columns
## [4] CNV gistic: RangedSummarizedExperiment with 5 rows and 4 columns
## Features:
## experiments() - obtain the ExperimentList instance
## colData() - the primary/phenotype DataFrame
## sampleMap() - the sample availability DataFrame
## `$`, `[`, `[[` - extract colData columns, subset, or experiment
## *Format() - convert into a long or wide DataFrame
## assays() - convert ExperimentList to a SimpleList of matricessubsetByColumn(myMultiAssay, as.list(allsamples)) #equivalent## harmonizing input:
## removing 3 sampleMap rows with 'colname' not in colnames of experiments## A MultiAssayExperiment object of 4 listed
## experiments with user-defined names and respective classes.
## Containing an ExperimentList class object of length 4:
## [1] Affy: ExpressionSet with 2 rows and 4 columns
## [2] Methyl 450k: matrix with 2 rows and 2 columns
## [3] Mirna: matrix with 4 rows and 3 columns
## [4] CNV gistic: RangedSummarizedExperiment with 5 rows and 4 columns
## Features:
## experiments() - obtain the ExperimentList instance
## colData() - the primary/phenotype DataFrame
## sampleMap() - the sample availability DataFrame
## `$`, `[`, `[[` - extract colData columns, subset, or experiment
## *Format() - convert into a long or wide DataFrame
## assays() - convert ExperimentList to a SimpleList of matricesYou can select certain assays / experiments using subset, by providing a character, logical, or integer vector. An example using character:
myMultiAssay[, , c("Affy", "CNV gistic")]## A MultiAssayExperiment object of 2 listed
## experiments with user-defined names and respective classes.
## Containing an ExperimentList class object of length 2:
## [1] Affy: ExpressionSet with 2 rows and 4 columns
## [2] CNV gistic: RangedSummarizedExperiment with 5 rows and 4 columns
## Features:
## experiments() - obtain the ExperimentList instance
## colData() - the primary/phenotype DataFrame
## sampleMap() - the sample availability DataFrame
## `$`, `[`, `[[` - extract colData columns, subset, or experiment
## *Format() - convert into a long or wide DataFrame
## assays() - convert ExperimentList to a SimpleList of matricesYou can subset assays also using logical or integer vectors:
is.cnv = grepl("CNV", names(experiments(myMultiAssay)))
is.cnv## [1] FALSE FALSE FALSE TRUEmyMultiAssay[, , is.cnv] #logical subsetting## A MultiAssayExperiment object of 1 listed
## experiment with a user-defined name and respective class.
## Containing an ExperimentList class object of length 1:
## [1] CNV gistic: RangedSummarizedExperiment with 5 rows and 4 columns
## Features:
## experiments() - obtain the ExperimentList instance
## colData() - the primary/phenotype DataFrame
## sampleMap() - the sample availability DataFrame
## `$`, `[`, `[[` - extract colData columns, subset, or experiment
## *Format() - convert into a long or wide DataFrame
## assays() - convert ExperimentList to a SimpleList of matricesmyMultiAssay[, , which(is.cnv)] #integer subsetting## A MultiAssayExperiment object of 1 listed
## experiment with a user-defined name and respective class.
## Containing an ExperimentList class object of length 1:
## [1] CNV gistic: RangedSummarizedExperiment with 5 rows and 4 columns
## Features:
## experiments() - obtain the ExperimentList instance
## colData() - the primary/phenotype DataFrame
## sampleMap() - the sample availability DataFrame
## `$`, `[`, `[[` - extract colData columns, subset, or experiment
## *Format() - convert into a long or wide DataFrame
## assays() - convert ExperimentList to a SimpleList of matricesRows of the assays correspond to assay features or measurements, such as genes. Regardless of whether the assay is ID-based (e.g., matrix, ExpressionSet) or range-based (e.g., RangedSummarizedExperiment), they can be subset using any of the following:
a character vector of IDs that will be matched to rownames in each assay
an integer vector that will select rows of this position from each assay. This probably doesn’t make sense unless every ExperimentList element represents the same measurements in the same order and will generate an error if any of the integer elements exceeds the number of rows in any ExperimentList element. The most likely use of integer subsetting would be as a head function, for example to look at the first 6 rows of each assay.
a logical vector that will be passed directly to the row subsetting operation for each assay.
a list or CharacterList of the same length as ExperimentList. Each element of the subsetting list will be passed on exactly to subset rows of the corresponding element of the ExperimentList.
Again, this operation always returns a MultiAssayExperiment class, unless “drop=TRUE” is passed to the [ backet subset, with any ExperimentList element not containing the feature having zero rows.
For example, return a MultiAssayExperiment where Affy and Methyl 450k contain only “ENST0000035076”" row, and “Mirna” and “CNV gistic” have zero rows (drop argument is set to FALSE by default in subsetBy*):
featSubsetted0 <- subsetByRow(myMultiAssay, "ENST00000355076")
featSubsetted0 <- myMultiAssay["ENST00000355076", , ] #equivalent
class(featSubsetted0)## [1] "MultiAssayExperiment"
## attr(,"package")
## [1] "MultiAssayExperiment"class(experiments(featSubsetted0))## [1] "ExperimentList"
## attr(,"package")
## [1] "MultiAssayExperiment"experiments(featSubsetted0)## ExperimentList class object of length 2:
## [1] Affy: ExpressionSet with 1 rows and 4 columns
## [2] Methyl 450k: matrix with 1 rows and 5 columnsIn the following, Affy ExpressionSet keeps both rows but with their order reversed, and Methyl 450k keeps only its second row.
featSubsetted <-
subsetByRow(myMultiAssay, c("ENST00000355076", "ENST00000294241"))
exprs(experiments(myMultiAssay)[[1]])## array1 array2 array3 array4
## ENST00000294241 101 103 105 107
## ENST00000355076 102 104 106 108exprs(experiments(featSubsetted)[[1]])## array1 array2 array3 array4
## ENST00000355076 102 104 106 108
## ENST00000294241 101 103 105 107GenomicRangesFor MultiAssayExperiment objects containing range-based objects (currently RangedSummarizedExperiment), these can be subset using a GRanges object, for example:
gr <- GRanges(seqnames = c("chr1", "chr1", "chr2"), strand = c("+", "+", "+"),
ranges = IRanges(start = c(330602, 443625, 934533),
end = c(330701, 443724, 934632)))Now do the subsetting. The function doing the work here is IRanges::subsetByOverlaps - see its arguments for flexible types of subsetting by range. The first three arguments here are for subset, the rest passed on to IRanges::subsetByOverlaps through “…”:
subsetted <- subsetByRow(myMultiAssay, gr, maxgap = 2L, type = "within")
experiments(subsetted)## ExperimentList class object of length 4:
## [1] Affy: ExpressionSet with 0 rows and 4 columns
## [2] Methyl 450k: matrix with 0 rows and 5 columns
## [3] Mirna: matrix with 0 rows and 3 columns
## [4] CNV gistic: RangedSummarizedExperiment with 0 rows and 4 columnsrowRanges(subsetted[[4]])## GRanges object with 0 ranges and 1 metadata column:
## seqnames ranges strand | feature_id
## <Rle> <IRanges> <Rle> | <character>
## -------
## seqinfo: 2 sequences from an unspecified genome; no seqlengthsSquare bracket subsetting can still be used here, but passing on arguments to IRanges::subsetByOverlaps through “…” is simpler using subsetByRow().
subsetByRow, subsetByColumn, subsetByAssay, and square bracket subsetting are all “endomorphic” operations, in that they always returns another MultiAssayExperiment object.
A double-bracket subset operation refers to an experiment, and will return the object contained within an ExperimentList element. It is not endomorphic. For example, the first ExperimentList element is called “Affy” and contains an ExpressionSet:
names(myMultiAssay)## [1] "Affy" "Methyl 450k" "Mirna" "CNV gistic"myMultiAssay[[1]]## ExpressionSet (storageMode: lockedEnvironment)
## assayData: 2 features, 4 samples
## element names: exprs
## protocolData: none
## phenoData
## sampleNames: array1 array2 array3 array4
## varLabels: slope53
## varMetadata: labelDescription
## featureData: none
## experimentData: use 'experimentData(object)'
## Annotation:myMultiAssay[["Affy"]]## ExpressionSet (storageMode: lockedEnvironment)
## assayData: 2 features, 4 samples
## element names: exprs
## protocolData: none
## phenoData
## sampleNames: array1 array2 array3 array4
## varLabels: slope53
## varMetadata: labelDescription
## featureData: none
## experimentData: use 'experimentData(object)'
## Annotation:complete.casesThe complete.cases function returns a logical vector identifying which primary units have data for all experiments. Recall that myMultiAssay provides data for four individuals:
colData(myMultiAssay)## DataFrame with 4 rows and 2 columns
## sex age
## <factor> <integer>
## Jack M 38
## Jill F 39
## Bob M 40
## Barbara F 41Of these, only Jack has data for all 5 experiments:
complete.cases(myMultiAssay)## [1] TRUE FALSE TRUE TRUEBut all four have coplete cases for Affy and Methyl 450k:
complete.cases(myMultiAssay[, , 1:2])## [1] TRUE TRUE TRUE TRUEThis output can be used to select individuals with complete data:
myMultiAssay[, complete.cases(myMultiAssay), ]## harmonizing input:
## removing 3 sampleMap rows with 'colname' not in colnames of experiments
## removing 1 colData rownames not in sampleMap 'primary'## A MultiAssayExperiment object of 4 listed
## experiments with user-defined names and respective classes.
## Containing an ExperimentList class object of length 4:
## [1] Affy: ExpressionSet with 2 rows and 3 columns
## [2] Methyl 450k: matrix with 2 rows and 4 columns
## [3] Mirna: matrix with 4 rows and 3 columns
## [4] CNV gistic: RangedSummarizedExperiment with 5 rows and 3 columns
## Features:
## experiments() - obtain the ExperimentList instance
## colData() - the primary/phenotype DataFrame
## sampleMap() - the sample availability DataFrame
## `$`, `[`, `[[` - extract colData columns, subset, or experiment
## *Format() - convert into a long or wide DataFrame
## assays() - convert ExperimentList to a SimpleList of matricesduplicatedThe duplicated function returns a list of LogicalLists, identifying replicate samples for each individual within each assay:
duplicated(myMultiAssay)## $Affy
## LogicalList of length 0
##
## $`Methyl 450k`
## LogicalList of length 1
## [["Jack"]] TRUE TRUE FALSE FALSE FALSE
##
## $Mirna
## LogicalList of length 0
##
## $`CNV gistic`
## LogicalList of length 0cThe combine c function allows the user to append an experiment to the list of experiments already present in MultiAssayExperiment. In the case that additional observations on the same set of samples were performed, the c function can conveniently be referenced to an existing assay that contains the same ordering of sample measurements.
Here we introduce a toy dataset created on the fly:
c(myMultiAssay, ExpScores = matrix(1:8, ncol = 4,
dim = list(c("ENSMBL0001", "ENSMBL0002"), paste0("pt", 1:4))),
mapFrom = 1L)## Warning in .local(x, ...): Assuming column order in the data provided
## matches the order in 'mapFrom' experiment(s) colnames## A MultiAssayExperiment object of 5 listed
## experiments with user-defined names and respective classes.
## Containing an ExperimentList class object of length 5:
## [1] Affy: ExpressionSet with 2 rows and 4 columns
## [2] Methyl 450k: matrix with 2 rows and 5 columns
## [3] Mirna: matrix with 4 rows and 3 columns
## [4] CNV gistic: RangedSummarizedExperiment with 5 rows and 4 columns
## [5] ExpScores: matrix with 2 rows and 4 columns
## Features:
## experiments() - obtain the ExperimentList instance
## colData() - the primary/phenotype DataFrame
## sampleMap() - the sample availability DataFrame
## `$`, `[`, `[[` - extract colData columns, subset, or experiment
## *Format() - convert into a long or wide DataFrame
## assays() - convert ExperimentList to a SimpleList of matricesNote: Alternatively, a sampleMap for the additional dataset can be provided.
MultiAssayExperimentA set of helper functions are available for clean-up operations. These include intersectRows, intersectColumns,mergeReplicates, complete.cases, and duplicated.
The intersectRows function takes all common rownames across experiments and returns a MultiAssayExperiment with those rows.
(noMatches <- intersectRows(myMultiAssay))## A MultiAssayExperiment object of 4 listed
## experiments with user-defined names and respective classes.
## Containing an ExperimentList class object of length 4:
## [1] Affy: ExpressionSet with 0 rows and 4 columns
## [2] Methyl 450k: matrix with 0 rows and 5 columns
## [3] Mirna: matrix with 0 rows and 3 columns
## [4] CNV gistic: RangedSummarizedExperiment with 0 rows and 4 columns
## Features:
## experiments() - obtain the ExperimentList instance
## colData() - the primary/phenotype DataFrame
## sampleMap() - the sample availability DataFrame
## `$`, `[`, `[[` - extract colData columns, subset, or experiment
## *Format() - convert into a long or wide DataFrame
## assays() - convert ExperimentList to a SimpleList of matricescomplete.cases returns a logical vector of colData rows that correspond to biological units that have data in all elements of the ExperimentList
complete.cases(myMultiAssay)## [1] TRUE FALSE TRUE TRUEcolData(myMultiAssay)[complete.cases(myMultiAssay), ]## DataFrame with 3 rows and 2 columns
## sex age
## <factor> <integer>
## Jack M 38
## Bob M 40
## Barbara F 41A call to intersectColumns returns another MultiAssayExperiment where the columns of each element of the ExperimentList correspond exactly to the rows of colData.
intersectColumns(myMultiAssay)## harmonizing input:
## removing 3 sampleMap rows with 'colname' not in colnames of experiments
## removing 1 colData rownames not in sampleMap 'primary'## A MultiAssayExperiment object of 4 listed
## experiments with user-defined names and respective classes.
## Containing an ExperimentList class object of length 4:
## [1] Affy: ExpressionSet with 2 rows and 3 columns
## [2] Methyl 450k: matrix with 2 rows and 4 columns
## [3] Mirna: matrix with 4 rows and 3 columns
## [4] CNV gistic: RangedSummarizedExperiment with 5 rows and 3 columns
## Features:
## experiments() - obtain the ExperimentList instance
## colData() - the primary/phenotype DataFrame
## sampleMap() - the sample availability DataFrame
## `$`, `[`, `[[` - extract colData columns, subset, or experiment
## *Format() - convert into a long or wide DataFrame
## assays() - convert ExperimentList to a SimpleList of matricesduplicated identifies primary column values or biological units that have multiple observations per assay. It returns a list of LogicalLists that indicate what biological units have one or more duplicated measurements. This output is used for merging replicates by default.
duplicated(myMultiAssay)## $Affy
## LogicalList of length 0
##
## $`Methyl 450k`
## LogicalList of length 1
## [["Jack"]] TRUE TRUE FALSE FALSE FALSE
##
## $Mirna
## LogicalList of length 0
##
## $`CNV gistic`
## LogicalList of length 0The mergeReplicates function allows the user to specify a function (default: mean) for combining replicate columns in each assay element. This can be combined with intersectColumns to create a MultiAssayExperiment object with one measurement in each experiment per biological unit.
mergeReplicates(intersectColumns(myMultiAssay))## harmonizing input:
## removing 3 sampleMap rows with 'colname' not in colnames of experiments
## removing 1 colData rownames not in sampleMap 'primary'## harmonizing input:
## removing 1 sampleMap rows with 'colname' not in colnames of experiments## A MultiAssayExperiment object of 4 listed
## experiments with user-defined names and respective classes.
## Containing an ExperimentList class object of length 4:
## [1] Affy: ExpressionSet with 2 rows and 3 columns
## [2] Methyl 450k: matrix with 2 rows and 3 columns
## [3] Mirna: matrix with 4 rows and 3 columns
## [4] CNV gistic: RangedSummarizedExperiment with 5 rows and 3 columns
## Features:
## experiments() - obtain the ExperimentList instance
## colData() - the primary/phenotype DataFrame
## sampleMap() - the sample availability DataFrame
## `$`, `[`, `[[` - extract colData columns, subset, or experiment
## *Format() - convert into a long or wide DataFrame
## assays() - convert ExperimentList to a SimpleList of matricesExtractor functions convert a MultiAssayExperiment into other forms that are convenient for analyzing. These would normally be called after any desired subsetting has been performed.
longFormat & wideFormatProduces long (default) or wide DataFrame objects. The following produces a long DataFrame (the default) for the first two assays:
longFormat(myMultiAssay[, , 1:2])## DataFrame with 18 rows and 5 columns
## assay primary rowname colname value
## <Rle> <Rle> <character> <Rle> <integer>
## 1 Affy Jack ENST00000294241 array1 101
## 2 Affy Jack ENST00000355076 array1 102
## 3 Affy Jill ENST00000294241 array2 103
## 4 Affy Jill ENST00000355076 array2 104
## 5 Affy Barbara ENST00000294241 array3 105
## ... ... ... ... ... ...
## 14 Methyl 450k Jill ENST00000383706 methyl3 6
## 15 Methyl 450k Barbara ENST00000355076 methyl4 7
## 16 Methyl 450k Barbara ENST00000383706 methyl4 8
## 17 Methyl 450k Bob ENST00000355076 methyl5 9
## 18 Methyl 450k Bob ENST00000383706 methyl5 10This is especially useful for performing regression against patient or sample data from colData using the pDataCols argument:
longFormat(myMultiAssay[, , 1:2], colDataCols="age")## DataFrame with 18 rows and 6 columns
## assay primary rowname colname value age
## <Rle> <Rle> <character> <Rle> <integer> <integer>
## 1 Affy Jack ENST00000294241 array1 101 38
## 2 Affy Jack ENST00000355076 array1 102 38
## 3 Affy Jill ENST00000294241 array2 103 39
## 4 Affy Jill ENST00000355076 array2 104 39
## 5 Affy Barbara ENST00000294241 array3 105 41
## ... ... ... ... ... ... ...
## 14 Methyl 450k Jill ENST00000383706 methyl3 6 39
## 15 Methyl 450k Barbara ENST00000355076 methyl4 7 41
## 16 Methyl 450k Barbara ENST00000383706 methyl4 8 41
## 17 Methyl 450k Bob ENST00000355076 methyl5 9 40
## 18 Methyl 450k Bob ENST00000383706 methyl5 10 40The “wide” format is useful for calculating correlations or performing regression against different genomic features. Wide format is in general not possible with replicate measurements, so we demonstrate on the cleaned MultiAssayExperiment for the first 5 columns:
maemerge <- mergeReplicates(intersectColumns(myMultiAssay))## harmonizing input:
## removing 3 sampleMap rows with 'colname' not in colnames of experiments
## removing 1 colData rownames not in sampleMap 'primary'## harmonizing input:
## removing 1 sampleMap rows with 'colname' not in colnames of experimentswideFormat(maemerge, colDataCols="sex")[, 1:5]## DataFrame with 3 rows and 5 columns
## primary sex Affy_ENST00000294241 Affy_ENST00000355076 CNV.gistic_a
## <factor> <factor> <numeric> <numeric> <numeric>
## 1 Jack M 101 102 7800.229
## 2 Bob M 107 108 2803.923
## 3 Barbara F 105 106 5455.331assay / assaysThe assay (singular) function takes a particular experiment and returns a matrix. By default, it will return the first experiment as a matrix.
assay(myMultiAssay)## array1 array2 array3 array4
## ENST00000294241 101 103 105 107
## ENST00000355076 102 104 106 108The assays (plural) function returns a SimpleList of data matrices from the ExperimentList:
assays(myMultiAssay)## List of length 4
## names(4): Affy Methyl 450k Mirna CNV gisticrownames and colnamesrownames and colnames return a CharacterList of row names and column names across all the assays. A CharacterList is an efficient alternative to list used when each element contains a character vector. It also provides a nice show method:
rownames(myMultiAssay)## CharacterList of length 4
## [["Affy"]] ENST00000294241 ENST00000355076
## [["Methyl 450k"]] ENST00000355076 ENST00000383706
## [["Mirna"]] hsa-miR-21 hsa-miR-191 hsa-miR-148a hsa-miR148b
## [["CNV gistic"]] a b c d ecolnames(myMultiAssay)## CharacterList of length 4
## [["Affy"]] array1 array2 array3 array4
## [["Methyl 450k"]] methyl1 methyl2 methyl3 methyl4 methyl5
## [["Mirna"]] micro1 micro2 micro3
## [["CNV gistic"]] mysnparray1 mysnparray2 mysnparray3 mysnparray4Any data classes in the ExperimentList object must support the following methods:
dimnames[dim()Here is what happens if one of the methods doesn’t:
objlist2 <- objlist
objlist2[[2]] <- as.vector(objlist2[[2]])
invalid.obj <- try(MultiAssayExperiment(objlist2, patient.data, dfmap))
invalid.obj## [1] "Error in if (dim(object)[2] > 0 && is.null(colnames(object))) { : \n missing value where TRUE/FALSE needed\n"
## attr(,"class")
## [1] "try-error"
## attr(,"condition")
## <simpleError in if (dim(object)[2] > 0 && is.null(colnames(object))) { paste(" colnames in", class(object), "are NULL")} else { NULL}: missing value where TRUE/FALSE needed>For more information on the formal API of MultiAssayExperiment, please see the API wiki document on GitHub. From within an R session, type:
API()to produce the webpage for the API or use the “shiny” argument to get a shiny app that browses the package, such as:
API(shiny=TRUE)The following methods are defined for MultiAssayExperiment:
methods(class="MultiAssayExperiment")## [1] $ $<- [ [[
## [5] [[<- assay assays c
## [9] coerce colData colData<- complete.cases
## [13] dimnames duplicated experiments experiments<-
## [17] isEmpty length longFormat mergeReplicates
## [21] metadata metadata<- names pData
## [25] pData<- sampleMap sampleMap<- show
## [29] subsetByAssay subsetByColData subsetByColumn subsetByRow
## [33] updateObject wideFormat
## see '?methods' for accessing help and source codec() function for adding new assays to existing MultiAssayExperiment
sessionInfo()## R version 3.4.2 (2017-09-28)
## Platform: x86_64-pc-linux-gnu (64-bit)
## Running under: Ubuntu 16.04.3 LTS
##
## Matrix products: default
## BLAS: /home/biocbuild/bbs-3.6-bioc/R/lib/libRblas.so
## LAPACK: /home/biocbuild/bbs-3.6-bioc/R/lib/libRlapack.so
##
## locale:
## [1] LC_CTYPE=en_US.UTF-8 LC_NUMERIC=C
## [3] LC_TIME=en_US.UTF-8 LC_COLLATE=C
## [5] LC_MONETARY=en_US.UTF-8 LC_MESSAGES=en_US.UTF-8
## [7] LC_PAPER=en_US.UTF-8 LC_NAME=C
## [9] LC_ADDRESS=C LC_TELEPHONE=C
## [11] LC_MEASUREMENT=en_US.UTF-8 LC_IDENTIFICATION=C
##
## attached base packages:
## [1] parallel stats4 stats graphics grDevices utils datasets
## [8] methods base
##
## other attached packages:
## [1] RaggedExperiment_1.2.0 SummarizedExperiment_1.8.0
## [3] DelayedArray_0.4.0 matrixStats_0.52.2
## [5] Biobase_2.38.0 GenomicRanges_1.30.0
## [7] GenomeInfoDb_1.14.0 IRanges_2.12.0
## [9] S4Vectors_0.16.0 BiocGenerics_0.24.0
## [11] MultiAssayExperiment_1.4.0 BiocStyle_2.6.0
##
## loaded via a namespace (and not attached):
## [1] tidyselect_0.2.2 reshape2_1.4.2 purrr_0.2.4
## [4] lattice_0.20-35 htmltools_0.3.6 yaml_2.1.14
## [7] rlang_0.1.2 R.oo_1.21.0 glue_1.2.0
## [10] R.utils_2.5.0 bindrcpp_0.2 bindr_0.1
## [13] GenomeInfoDbData_0.99.1 R.cache_0.12.0 plyr_1.8.4
## [16] stringr_1.2.0 zlibbioc_1.24.0 R.methodsS3_1.7.1
## [19] evaluate_0.10.1 knitr_1.17 httpuv_1.3.5
## [22] Rcpp_0.12.13 xtable_1.8-2 backports_1.1.1
## [25] XVector_0.18.0 mime_0.5 R.rsp_0.41.0
## [28] digest_0.6.12 stringi_1.1.5 bookdown_0.5
## [31] dplyr_0.7.4 shiny_1.0.5 grid_3.4.2
## [34] rprojroot_1.2 tools_3.4.2 bitops_1.0-6
## [37] magrittr_1.5 RCurl_1.95-4.8 tibble_1.3.4
## [40] pkgconfig_2.0.1 tidyr_0.7.2 Matrix_1.2-11
## [43] shinydashboard_0.6.1 assertthat_0.2.0 rmarkdown_1.6
## [46] R6_2.2.2 compiler_3.4.2