Package 'SEtools'

aggSE

Description

Aggregates the rows of a 'SummarizedExperiment'.

Usage

aggSE(x, by, assayFun = NULL, rowDatFuns = list())

Arguments

x	An object of class 'SummarizedExperiment'
by	Vector by which to aggregate, or column of 'rowData(x)'
assayFun	Function by which to aggregate, or a list of such functions (or vector of function names) of the same length as there are assays. If NULL will attempt to use an appropriate function (and notify the functions used), typically the mean.
rowDatFuns	A named list providing functions by which to aggregate each rowData columns. If a given column has no specified function, the default will be used, i.e. logical are transformed into a proportion, numerics are aggregated by median, and unique factors/characters are pasted together. Use 'rowDataFuns=NULL' to discard rowData.

Value

An object of class 'SummarizedExperiment'

Examples

library(SummarizedExperiment)
data("SE", package="SEtools")
# arbitrary IDs for example aggregation:
rowData(SE)$otherID <- rep(LETTERS[1:10],each=10)
SE <- aggSE(SE, "otherID")

---

castSE

Description

Casts a data.frame as a SummarizedExperiment-class

Usage

castSE(
  x,
  rowNames = NULL,
  colNames = NULL,
  assayNames = NULL,
  colData = NULL,
  rowData = NULL,
  sparse = FALSE
)

Arguments

x	A data.frame
rowNames	Column of 'x' containing the row.names (if omitted, will build from 'rowData')
colNames	Column of 'x' containing the column names (if omitted, will build from 'colData')
assayNames	Columns of 'x' to turn into assays
colData	Columns of 'x' to use as colData
rowData	Columns of 'x' to use as rowData
sparse	Local, whether to keep the assays sparse.

Value

A SummarizedExperiment-class

Examples

d <- data.frame(transcript=rep(LETTERS[1:10],each=2), gene=rep(LETTERS[1:5],each=4),
                count=rpois(20, 10), sample=letters[1:2])
head(d)
castSE(d, rowData=c("transcript","gene"), colNames="sample")

---

Example dataset

Description

A SummarizedExperiment-class containing (a subset of) whole-hippocampus RNAseq of mice after different stressors.

Value

a SummarizedExperiment-class.

References

Floriou-Servou et al. (2018). Distinct Proteomic, Transcriptomic, and Epigenetic Stress Responses in Dorsal and Ventral Hippocampus. Biological Psychiatry, 84(7): 531-541. DOI: 10.1016/j.biopsych.2018.02.003.

---

flattenPB

Description

Flattens a pseudo-bulk SummarizedExperiment as produced by 'muscat::aggregateData' so that all cell types are represented in a single assay. Optionally normalizes the data and calculates per-sample logFCs.

Usage

flattenPB(pb, norm = TRUE, lfc_group = NULL)

Arguments

pb	a pseudo-bulk SummarizedExperiment as produced by 'muscat::aggregateData', with different celltypes/clusters are assays.
norm	Logical; whether to calculate logcpm (TMM normalization).
lfc_group	the colData column to use to calculate foldchange. If NULL (default), no foldchange assay will be computed.

Value

A SummarizedExperiment

---

log2FC

Description

Generates log2(foldchange) matrix/assay, eventually on a per-batch fashion.

Usage

log2FC(
  x,
  fromAssay = NULL,
  controls,
  by = NULL,
  isLog = NULL,
  agFun = rowMeans,
  toAssay = "log2FC"
)

Arguments

x	A numeric matrix, or a 'SummarizedExperiment' object
fromAssay	The assay to use if 'x' is a 'SummarizedExperiment'
controls	A vector of which samples should be used as controls for foldchange calculations.
by	An optional vector indicating groups/batches by which the controls will be averaged to calculate per-group foldchanges.
isLog	Logical; whether the data is log-transformed. If NULL, will attempt to figure it out from the data and/or assay name
agFun	Aggregation function for the baseline (default rowMeans)
toAssay	The name of the assay in which to save the output.

Value

An object of same class as 'x'; if a 'SummarizedExperiment', will have the additional assay named from 'toAssay'.

Examples

log2FC( matrix(rnorm(40), ncol=4), controls=1:2 )

---

mergeSEs

Description

Merges a list of SummarizedExperiment-class, either by row.names or through specified rowData fields. In cases of many-to-many (or one-to-many) mappings, 'aggFun' determines whether the records are aggregated by linking ID (if an aggregation method is given) or all combinations are returned (if 'aggFun=NULL' - default).

Usage

mergeSEs(
  ll,
  use.assays = NULL,
  do.scale = TRUE,
  commonOnly = TRUE,
  colColumns = NULL,
  mergeBy = NULL,
  aggFun = NULL,
  addDatasetPrefix = TRUE,
  defValues = list(),
  keepRowData = TRUE,
  BPPARAM = SerialParam()
)

Arguments

ll	A (named) list of SummarizedExperiment-class
use.assays	Names (or indexes) of the assays to use. By default, all common assays are used.
do.scale	A logical vector indicating (globally or for each assay) whether to perform row unit-variance scaling on each dataset before merging (default TRUE).
commonOnly	Logical; whether to restrict to rows present in all datasets (default TRUE).
colColumns	A character vector specifying 'colData' columns to include (if available in at least one of the datasets). If NULL, everything is kept.
mergeBy	The 'rowData' column to merge with. If NULL, row.names are used.
aggFun	The aggregation function to use when multiple rows have the same 'mergeBy' value. If merging multiple assays, a different function per assay can be passed as a named list (see aggSE). If NULL (default), entries will be reused to have each combination.
addDatasetPrefix	Logical; whether the name of the dataset should be appended to the sample names (default TRUE).
defValues	An optional named list of default 'colData' values when some columns are missing from some SEs.
keepRowData	Logical, whether to keep the rowData (default TRUE).
BPPARAM	For multithreading the aggregation step.

Value

An object of class SummarizedExperiment-class

Examples

data("SE", package="SEtools")
mergeSEs( list( se1=SE[,1:10], se2=SE[,11:20] ) )

---

resetAllSEtoolsOptions

Description

Resents all global options relative to SEtools.

Usage

resetAllSEtoolsOptions()

Value

None

Examples

resetAllSEtoolsOptions()

---

se2xlsx

Description

Writes a SummarizedExperiment to an excel/xlsx file. Requires the 'openxlsx' package.

Usage

se2xls(se, filename, addSheets = NULL)

Arguments

se	The 'SummarizedExperiment'
filename	xlsx file name
addSheets	An optional list of additional tables to save as sheets.

Value

Saves to file.

Examples

data("SE", package="SEtools")
# not run
# se2xls(SE, filename="SE.xlsx")

---

sehm

Description

Deprecated pheatmap wrapper for SummarizedExperiment-class. **This function has been replaced by the sechm function from the 'sechm' package and is retained here solely for backward compatibility.**

Usage

sehm(
  se,
  genes,
  do.scale = FALSE,
  assayName = .getDef("assayName"),
  sortRowsOn = seq_len(ncol(se)),
  cluster_cols = FALSE,
  cluster_rows = is.null(sortRowsOn),
  toporder = NULL,
  hmcols = NULL,
  breaks = .getDef("breaks"),
  gaps_at = .getDef("gaps_at"),
  gaps_row = NULL,
  anno_rows = .getDef("anno_rows"),
  anno_columns = .getDef("anno_columns"),
  anno_colors = NULL,
  show_rownames = NULL,
  show_colnames = FALSE,
  ...
)

Arguments

se	A SummarizedExperiment-class.
genes	An optional vector of genes (i.e. row names of 'se')
do.scale	Logical; whether to scale rows (default FALSE).
assayName	An optional vector of assayNames to use. The first available will be used, or the first assay if NULL.
sortRowsOn	Sort rows by MDS polar order using the specified columns (default all)
cluster_cols	Whether to cluster columns (default F)
cluster_rows	Whether to cluster rows; default FALSE if 'do.sortRows=TRUE'.
toporder	Optional verctor of categories on which to supra-order when sorting rows, or name of a 'rowData' column to use for this purpose.
hmcols	Colors for the heatmap.
breaks	Breaks for the heatmap colors. Alternatively, symmetrical breaks can be generated automatically by setting 'breaks' to a numerical value between 0 and 1. The value is passed as the 'split.prop' argument to the getBreaks function, and indicates the proportion of the points to map to a linear scale, while the more extreme values will be plotted on a quantile scale. 'breaks=FALSE' will disable symmetrical scale and quantile capping, while retaining automatic breaks. 'breaks=1' will produce a symmetrical scale without quantile capping.
gaps_at	Columns of 'colData' to use to establish gaps between columns.
gaps_row	Passed to the heatmap function; if missing, will be set automatically according to toporder.
anno_rows	Columns of 'rowData' to use for left annotation.
anno_columns	Columns of 'colData' to use for top annotation.
anno_colors	List of colors to use for annotation.
show_rownames	Whether to show row names (default TRUE if less than 50 rows to plot).
show_colnames	Whether to show column names (default FALSE).
...	Further arguments passed to 'pheatmap'

Value

A heatmap.

---

svacor

Description

A wrapper around SVA-based correction, providing a corrected assay. If this is RNAseq data or similar, use a count assay assay with 'useVST=TRUE'; otherwise (e.g. proteomics) a log-normalized assay is recommended.

Usage

svacor(
  SE,
  form,
  form0 = ~1,
  assayName = NULL,
  regressOutNull = TRUE,
  useVST = TRUE,
  n.sv = NULL,
  ...
)

Arguments

SE	An object of class 'SummarizedExperiment'.
form	The formula of the differential expression model
form0	An optional formula for the null model
assayName	The name (or index) of the assay to use.
regressOutNull	Logical; whether to regress out the variables of 'form0'.
useVST	Logical; whether to use DESeq2's variance-stabilizing transformation; (for count data!)
n.sv	The number of surrogate variables (if omitted, sva will attempt to estimate it)
...	Any other argument passed to the sva command.

Value

Returns the 'SummarizedExperiment' with a 'corrrected' assay and the surrogate variables in 'colData'.

Examples

data("SE", package="SEtools")
SE <- svacor(SE, ~Condition)

---