get_reducedDimPlot.sce: Visualizing tSNE and PCA results
In evanbiederstedt/sandbox: The functions

Description Usage Arguments Details Value See Also

This will generate a ggplot starting from a DimRedPlot object. See get_reducedDimPlot.sce to start from a SingleCellExperiment object.

Wrapper function for extracting the X and Y coordinates for the plot as well as the variables for coloring, shape assignment etc. This function expects a SingleCellExperiment object.

plt.DimRedPlot(drp_object, X = "x_axs", Y = "y_axs", color_by = NULL,
  shape_by = NULL, size_by = NULL, circle_by = NULL, which_quantile = 1,
  ignore_drp_labels = FALSE, theme_size = 14, legend = "auto",
  alpha = 0.75, remove_rug = FALSE, set_colors = TRUE,
  set_fill_colors = TRUE)

get_reducedDimPlot.sce(object, which_reddim = NULL, which_pcs = c(1, 2),
  X = NULL, Y = NULL, dim_red_type = NULL, color_by = NULL,
  shape_by = NULL, size_by = NULL, circle_by = NULL,
  exprs_values = "counts", add_cell_info = NULL)

`drp_object`	a `DimRedPlot` object
`X`	name of the drp_object column that contains the values for the x axis
`Y`	name of the drp_object column that contains the values for the y axis
`color_by`	specify the column of `drp_object$plot_data` whose entries will be used to assign either discrete or continous color schemes.
`shape_by`	the factor to assign the shape of the points to; max. 10 levels are accepted for `shape_by`. If you want to specify the (same) shape for all points, use an integer (default shape is 19).
`size_by`	the factor to assign the size of the points to; default: `NULL`. If you want to specify the (same) size for all points, use one integer (the default size is 3).
`circle_by`	a factor (e.g. "condition") that will be used to draw circles around the points belonging to the same factor instance.
`which_quantile`	default: 1
`ignore_drp_labels`	If set to TRUE, all entries of `drp_object$factors` will be ignored. If you want to ignore only selected labels, indicate their name (e.g. c("color_by", "exprs_val_type")). The default setting (`FALSE`) will make use of the information from the drp_object unless specified here via `color_by`, `shape_by`, `size_by`, `circle_by`.
`theme_size`	font size for the plot
`legend`	choose wether there should be no legend ("none"), a legend for every factor or whether the legend will only show up for those factors that have more than 1 level ("auto", default setting).
`alpha`	define the opacity of the points, default: .65
`remove_rug`	Define whether the bars at both axes that represent the two 1d marginal distributions should be removed. Default: FALSE.
`set_colors`	Set to `FALSE` if you want to add your own color scheme to the resulting plot. The default behavior is to try to automatically assign an optimized coloring scheme. Default: `TRUE`.
`set_fill_colors`	Set to `FALSE` if you want to add your own color scheme to the resulting plot for the background circles. The default behavior is to try to automatically assign an optimized coloring scheme.
`object`	`SingleCellExperiment` object
`which_reddim`	accessor to retrieve the list of choice for the reduced dimension coordinates from `reducedDims(object)`. See `details`.
`which_pcs`	vector of integers to specify which (principal) components are going to be used for the x and y coordinates. The default (`c(1,2)`) will assign PC1 to the X axis and PC2 to the Y axis. This parameter will be ignored if X and Y are specified directly.
`dim_red_type`	what type of dimensionality reduction was used to generate the data (e.g. "PCA"). Will mostly be used for labelling purposes. Default: NULL
`exprs_values`	types of expression values that will be used for plotting if a gene name is being specified in either `color_by`, `shape_by`, `size_by`
`add_cell_info`	if you want to add columns for certain entries of `colData(object)`, specify their names here. Default: NULL
`X`	directly supply the X coordinates, e.g. `pca_res$PC1`; if this is used, `which_reddim` and `which_pcs` are ignored. `X` should be of the same length (and order!) as `colData(object)` (or `pData2.0(object)`).
`Y`	directly supply the Y coordinates, e.g. `pca_res$PC2`. Should be of the same length (and order!) as `colData(object)` (or `pData2.0(object)`). `which_reddim` and `which_pcs` will be ignored if `X` and `Y` are specified.
`color_by`	entry of either `colData(object)`, `rowData(object)`, `rownames(object)`, which will be used to assign either discrete or continous color schemes. Alternatively, you can supply a `list` with a label stored in `list$title` and values stored in `list$result` (`result` should be as long as `dim(object)[2]`. Default: `NULL`
`shape_by`	default: `NULL`
`size_by`	default: `NULL`
`circle_by`	Used for encircling points of interest; default: `NULL`

which_reddim The reducedDims slot of the SingleCellExperiment object is a list (SimpleList. To know which coordinates to extract, the accessor for the specific reducedDims list member must be given.

ggplot2 object (a plot)

DimRedPlot object: a simple S3-based object, i.e. a list with:

plot_data: data.frame fit for ggplot2-based plotting
label_data: additional stuff that may be used for labelling the x and y axes of the final plot
factors: names of the plotting variables, e.g. which values were used for defining the coloring etc.

get_reducedDimPlot.sce, which generates the DimRedPlot object that will be used here.

SingleCellExperiment, reducedDims, plt.DimRedPlot

evanbiederstedt/sandbox documentation built on May 26, 2019, 12:31 p.m.