layer_boxplot.Rd

% Generated by roxygen2: do not edit by hand
% Please edit documentation in R/layer_boxplot.R
\name{layer_boxplot}
\alias{layer_boxplot}
\title{Layer-level (group-level) boxplots}
\usage{
layer_boxplot(
  i = 1,
  sig_genes = sig_genes_extract(),
  short_title = TRUE,
  sce_layer = fetch_data(type = "sce_layer"),
  col_bkg_box = "grey80",
  col_bkg_point = "grey40",
  col_low_box = "violet",
  col_low_point = "darkviolet",
  col_high_box = "skyblue",
  col_high_point = "dodgerblue4",
  cex = 2,
  group_var = "layer_guess_reordered_short",
  assayname = "logcounts"
)
}
\arguments{
\item{i}{A \code{integer(1)} indicating which row of \code{sig_genes} do you want to
plot.}

\item{sig_genes}{The output of \code{\link[=sig_genes_extract_all]{sig_genes_extract_all()}}.}

\item{short_title}{A \code{logical(1)} indicating whether to print a short title
or not.}

\item{sce_layer}{Defaults to the output of
\code{fetch_data(type = 'sce_layer')}. This is a
\linkS4class{SingleCellExperiment}
object with the spot-level Visium data compressed via pseudo-bulking to the
layer-level (group-level) resolution. See \code{\link[=fetch_data]{fetch_data()}} for more details.}

\item{col_bkg_box}{Box background color for layers not used when visualizing
the \code{pairwise} model results.}

\item{col_bkg_point}{Similar to \code{col_bkg_box} but for the points.}

\item{col_low_box}{Box background color for layer(s) with the expected
lower expression based on the actual test for row \code{i} of \code{sig_genes}.}

\item{col_low_point}{Similar to \code{col_low_box} but for the points.}

\item{col_high_box}{Similar to \code{col_low_box} but for the expected layer(s)
with higher expression.}

\item{col_high_point}{Similar to \code{col_high_box} but for the points.}

\item{cex}{Controls the size of the text, points and axis legends.}

\item{group_var}{A \code{character(1)} specifying a \code{colData(sce_layer)} column
name to use for the x-axis.}

\item{assayname}{A \code{character(1)} specifying the default assay to use from
\code{assays(sce_layer)}.}
}
\value{
This function creates a boxplot of the layer-level data
(group-level) separated by layer and colored based on the model type from row
\code{i} of \code{sig_genes}.
}
\description{
This function uses the output of \code{\link[=sig_genes_extract_all]{sig_genes_extract_all()}} as well as the
logcounts from the layer-level (group-level) data to visualize the expression
of a given gene and display the modeling results for the given gene.
}
\examples{

## Obtain the necessary data
if (!exists("modeling_results")) {
    modeling_results <- fetch_data(type = "modeling_results")
}
if (!exists("sce_layer")) sce_layer <- fetch_data(type = "sce_layer")

## Top 2 genes from the enrichment model
sig_genes <- sig_genes_extract_all(
    n = 2,
    modeling_results = modeling_results,
    sce_layer = sce_layer
)

## Example default boxplot
set.seed(20200206)
layer_boxplot(sig_genes = sig_genes, sce_layer = sce_layer)

## Now show the long title version
set.seed(20200206)
layer_boxplot(
    sig_genes = sig_genes,
    short_title = FALSE,
    sce_layer = sce_layer
)

set.seed(20200206)
layer_boxplot(
    i = which(sig_genes$model_type == "anova")[1],
    sig_genes = sig_genes,
    sce_layer = sce_layer
)

set.seed(20200206)
layer_boxplot(
    i = which(sig_genes$model_type == "pairwise")[1],
    sig_genes = sig_genes,
    sce_layer = sce_layer
)

## Viridis colors displayed in the shiny app
library("viridisLite")
set.seed(20200206)
layer_boxplot(
    sig_genes = sig_genes,
    sce_layer = sce_layer,
    col_low_box = viridis(4)[2],
    col_low_point = viridis(4)[1],
    col_high_box = viridis(4)[3],
    col_high_point = viridis(4)[4]
)

## Paper colors displayed in the shiny app
set.seed(20200206)
layer_boxplot(
    sig_genes = sig_genes,
    sce_layer = sce_layer,
    col_low_box = "palegreen3",
    col_low_point = "springgreen2",
    col_high_box = "darkorange2",
    col_high_point = "orange1"
)

## Blue/red colors displayed in the shiny app
set.seed(20200206)
layer_boxplot(
    i = which(sig_genes$model_type == "pairwise")[1],
    sig_genes = sig_genes,
    sce_layer = sce_layer,
    col_bkg_box = "grey90",
    col_bkg_point = "grey60",
    col_low_box = "skyblue2",
    col_low_point = "royalblue3",
    col_high_box = "tomato2",
    col_high_point = "firebrick4",
    cex = 3
)
}
\references{
Adapted from
https://github.com/LieberInstitute/HumanPilot/blob/master/Analysis/Layer_Guesses/layer_specificity.R
}
\seealso{
Other Layer modeling functions: 
\code{\link{sig_genes_extract}()},
\code{\link{sig_genes_extract_all}()}
}
\concept{Layer modeling functions}