Generate correlation plots between predicted and expected cell type proportions from test data

Source: R/evalMetrics.R

Generate correlation plot between predicted and expected cell type proportions from test data. Correlation plots can be displayed all mixed or split by cell type (CellType) or number of cell types present in the samples (nCellTypes). See the facet.by argument and examples for more information. Moreover, a user-selected correlation value is displayed as an annotation on the plots. See the corr argument for details.

corrExpPredPlot(
  object,
  colors,
  facet.by = NULL,
  color.by = "CellType",
  corr = "both",
  filter.sc = TRUE,
  pos.x.label = 0.01,
  pos.y.label = 0.95,
  sep.labels = 0.15,
  size.point = 0.1,
  alpha.point = 1,
  ncol = NULL,
  nrow = NULL,
  title = NULL,
  theme = NULL,
  ...
)

Arguments

object

DigitalDLSorter object with trained.model slot containing metrics in the test.deconv.metrics slot of a DigitalDLSorterDNN object.

colors

Vector of colors to be used. Only vectors with a number of colors equal to or greater than the levels of color.by will be accepted. By default, a custom color list is used.

facet.by

Variable used to display data in different panels. If NULL, the plot is not split into different panels. Options are nCellTypes (by number of different cell types) and CellType (by cell type).

color.by

Variable used to color data. Options are nCellTypes and CellType.

corr

Correlation value displayed as an annotation on the plot. Available metrics are Pearson's correlation coefficient ('pearson') and concordance correlation coefficient ('ccc'). The argument can be 'pearson', 'ccc' or 'both' (by default).

filter.sc

Boolean indicating whether single-cell profiles are filtered out and only errors associated with pseudo-bulk samples are displayed (TRUE by default).

pos.x.label

X-axis position of correlation annotations (0.95 by default).

pos.y.label

Y-axis position of correlation annotations (0.1 by default).

sep.labels

Space separating annotations if corr is equal to 'both' (0.15 by default).

size.point

Size of points (0.1 by default).

alpha.point

Alpha of points (0.1 by default).

ncol

Number of columns if facet.by is other than NULL.

nrow

Number of rows if facet.by is different from NULL.

title

Title of the plot.

theme

ggplot2 theme.

...

Additional arguments for the facet_wrap function from ggplot2 if facet.by is not NULL.

Value

A ggplot object with the correlation plots between expected and actual proportions.

Examples

if (FALSE) { # \dontrun{
set.seed(123)
sce <- SingleCellExperiment::SingleCellExperiment(
  assays = list(
    counts = matrix(
      rpois(30, lambda = 5), nrow = 15, ncol = 20,
      dimnames = list(paste0("Gene", seq(15)), paste0("RHC", seq(20)))
    )
  ),
  colData = data.frame(
    Cell_ID = paste0("RHC", seq(20)),
    Cell_Type = sample(x = paste0("CellType", seq(6)), size = 20,
                       replace = TRUE)
  ),
  rowData = data.frame(
    Gene_ID = paste0("Gene", seq(15))
  )
)
DDLS <- createDDLSobject(
  sc.data = sce,
  sc.cell.ID.column = "Cell_ID",
  sc.gene.ID.column = "Gene_ID",
  sc.filt.genes.cluster = FALSE, 
  sc.log.FC = FALSE
)
probMatrixValid <- data.frame(
  Cell_Type = paste0("CellType", seq(6)),
  from = c(1, 1, 1, 15, 15, 30),
  to = c(15, 15, 30, 50, 50, 70)
)
DDLS <- generateBulkCellMatrix(
  object = DDLS,
  cell.ID.column = "Cell_ID",
  cell.type.column = "Cell_Type",
  prob.design = probMatrixValid,
  num.bulk.samples = 50,
  verbose = TRUE
)
# training of DDLS model
tensorflow::tf$compat$v1$disable_eager_execution()
DDLS <- trainDDLSModel(
  object = DDLS,
  on.the.fly = TRUE,
  batch.size = 15,
  num.epochs = 5
)
# evaluation using test data
DDLS <- calculateEvalMetrics(
  object = DDLS
)
# correlations by cell type
corrExpPredPlot(
  object = DDLS,
  facet.by = "CellType",
  color.by = "CellType",
  corr = "both"
)
# correlations of all samples mixed
corrExpPredPlot(
  object = DDLS,
  facet.by = NULL,
  color.by = "CellType",
  corr = "ccc",
  pos.x.label = 0.2,
  alpha.point = 0.3
)
} # }