SconeExperiment-class: Class SconeExperiment
In scone: Single Cell Overview of Normalized Expression data

Description Usage Arguments Details Value Slots See Also Examples

Objects of this class store, at minimum, a gene expression matrix and a set of covariates (sample metadata) useful for running scone. These include, the quality control (QC) metrics, batch information, and biological classes of interest (if available).

The typical way of creating SconeExperiment objects is via a call to the SconeExperiment function or to the scone function. If the object is a result to a scone call, it will contain the results, e.g., the performance metrics, scores, and normalization workflow comparisons. (See Slots for a full list).

This object extends the SummarizedExperiment class.

The constructor SconeExperiment creates an object of the class SconeExperiment.

SconeExperiment(object, ...)

## S4 method for signature 'SummarizedExperiment'
SconeExperiment(object,
  which_qc = integer(), which_bio = integer(),
  which_batch = integer(), which_negconruv = integer(),
  which_negconeval = integer(), which_poscon = integer(),
  is_log = FALSE)

## S4 method for signature 'matrix'
SconeExperiment(object, qc, bio, batch,
  negcon_ruv = NULL, negcon_eval = negcon_ruv, poscon = NULL,
  is_log = FALSE)

`object`	Either a matrix or a `SummarizedExperiment` containing the raw gene expression.
`...`	see specific S4 methods for additional arguments.
`which_qc`	index that specifies which columns of 'colData' correspond to QC measures.
`which_bio`	index that specifies which column of 'colData' corresponds to 'bio'.
`which_batch`	index that specifies which column of 'colData' corresponds to 'batch'.
`which_negconruv`	index that specifies which column of 'rowData' has information on negative controls for RUV.
`which_negconeval`	index that specifies which column of 'rowData' has information on negative controls for evaluation.
`which_poscon`	index that specifies which column of 'rowData' has information on positive controls.
`is_log`	are the expression data in log scale?
`qc`	numeric matrix with the QC measures.
`bio`	factor with the biological class of interest.
`batch`	factor with the batch information.
`negcon_ruv`	a logical vector indicating which genes to use as negative controls for RUV.
`negcon_eval`	a logical vector indicating which genes to use as negative controls for evaluation.
`poscon`	a logical vector indicating which genes to use as positive controls.

The QC matrix, biological class, and batch information are stored as elements of the 'colData' of the object.

The positive and negative control genes are stored as elements of the 'rowData' of the object.

A SconeExperiment object.

which_qc: integer. Index of columns of 'colData' that contain the QC metrics.
which_bio: integer. Index of the column of 'colData' that contains the biological classes information (it must be a factor).
which_batch: integer. Index of the column of 'colData' that contains the batch information (it must be a factor).
which_negconruv: integer. Index of the column of 'rowData' that contains a logical vector indicating which genes to use as negative controls to infer the factors of unwanted variation in RUV.
which_negconeval: integer. Index of the column of 'rowData' that contains a logical vector indicating which genes to use as negative controls to evaluate the performance of the normalizations.
which_poscon: integer. Index of the column of 'rowData' that contains a logical vector indicating which genes to use as positive controls to evaluate the performance of the normalizations.
hdf5_pointer: character. A string specifying to which file to write / read the normalized data.
imputation_fn: list of functions used by scone for the imputation step.
scaling_fn: list of functions used by scone for the scaling step.
scone_metrics: matrix. Matrix containing the "raw" performance metrics. See scone for a description of each metric.
scone_scores: matrix. Matrix containing the performance scores (transformed metrics). See scone for a discussion on the difference between scores and metrics.
scone_params: data.frame. A data frame containing the normalization schemes applied to the data and compared.
scone_run: character. Whether scone was run and in which mode ("no", "in_memory", "hdf5").
is_log: logical. Are the expression data in log scale?
nested: logical. Is batch nested within bio? (Automatically set by scone).
rezero: logical. TRUE if scone was run with zero="preadjust" or zero="strong".
fixzero: logical. TRUE if scone was run with zero="postadjust" or zero="strong".
impute_args: list. Arguments passed to all imputation functions.

get_normalized, get_params, get_batch, get_bio, get_design, get_negconeval, get_negconruv, get_poscon, get_qc, get_scores, and get_score_ranks to access internal fields, select_methods for subsetting by method, and scone for running scone workflows.

set.seed(42)
nrows <- 200
ncols <- 6
counts <- matrix(rpois(nrows * ncols, lambda=10), nrows)
rowdata <- data.frame(poscon=c(rep(TRUE, 10), rep(FALSE, nrows-10)))
coldata <- data.frame(bio=gl(2, 3))
se <- SummarizedExperiment(assays=SimpleList(counts=counts),
                          rowData=rowdata, colData=coldata)

scone1 <- SconeExperiment(assay(se), bio=coldata$bio, poscon=rowdata$poscon)

scone2 <- SconeExperiment(se, which_bio=1L, which_poscon=1L)