csTopTable: Compute Cell-Specific Statistics
In shenorrLab/csSAM: csSAM - cell-specific Significance Analysis of Microarrays

Description Usage Arguments Value Methods (by class) See Also

csTopTable is a generic function that returns cell type-level statistics on biological features, e.g., cell type-specific expression.

csTopTable(x, ...)

## Default S3 method:
csTopTable(x, n = 100L, sort.by = 1L,
  decreasing = FALSE, select = NULL, subset = names(threshold),
  threshold = NULL, annotation = NULL, ..., merge = FALSE,
  nodups = FALSE)

## S3 method for class 'array'
csTopTable(x, ...)

## S3 method for class 'character'
csTopTable(x, ...)

## S3 method for class 'csSAMfit'
csTopTable(x, alternative = c("two.sided", "greater",
  "less", "min"), sort.by = "FDR", decreasing = FALSE, ...)

`x`	data object, typically obtained by fitting a model that computes cell type-specific statistics, e.g., csSAMfit.
`...`	extra parameters passed to specific methods.
`n`	maximum number of top features to retain – in each cell type. If `NULL` or `Inf` then all selected features are returned. Otherwise the first `n` feature in each cell type are returned, after ordering according to arguments `sort.by` and `decreasing`.
`sort.by`	variable to use to order the features.
`decreasing`	logical that controls the feature ordering based on `sort.by`.
`select`	specifies the statistic to include in the result.
`subset`	indicates the subset of tissue/cell type to include in the result. Default is to return only the cell types for which a `threshold` value was provided. If `NULL`, as default, then no subsetting is performed.
`threshold`	numeric vector that indicates the threshold applied to select features based on their values of `sort.by`. Thresholds can be defined for specific cell types by using naming the elements of `threshold` with corresponding cell type names. The default threshold that is applied to all cell types is taken from the first unnamed element. Depending on `decreasing`, thresholds are interpreted differently: `decreasing = FALSE` keeps features with `sort.by` "<= x" `decreasing = TRUE` keeps features with `sort.by` ">= x"
`annotation`	Bioconductor annotation package to use to annotate the features (rownames) using xbioc::geneInfo.
`merge`	logical that indicates if the result should be a a list of `data.frame` objects (`merge = FALSE`), or a single `data.frame` object with columns "Feature" and "Cell type" containing the feature identifiers and cell type respectively.
`nodups`	a logical that indicates if features appearing in multiple cell types should be removed. This is useful in combination with `threshold` to extract features that are found significant in a single cell type.
`alternative`	Alternative hypothesis to consider. Alternative "min" returns, for each features, the alternative achieving the minimum FDR.

a list of data.frame objects if merge = FALSE, or a single data.frame object if merge = TRUE.

If merge=TRUE and select is of length one, then the result is a matrix object (feature x cell type) containing the selected value.

csSAMfit: returns, for each feature, the false discovery rates of differential expression between groups of samples within each cell type, as computed by fdrCsSAM when running csSAM. These are returned as a list, whith one element per cell type.

fdrCsSAM

shenorrLab/csSAM documentation built on May 29, 2019, 9:23 p.m.