plot_loadings_annot: Plot the gene by celltype loadings for a factor

In scITD: Single-Cell Interpretable Tensor Decomposition

Plot the gene by celltype loadings for a factor

Description

Plot the gene by celltype loadings for a factor

Usage

plot_loadings_annot(
  container,
  factor_select,
  use_sig_only = FALSE,
  nonsig_to_zero = FALSE,
  annot = "none",
  pathways = NULL,
  sim_de_donor_group = NULL,
  sig_thresh = 0.05,
  display_genes = FALSE,
  gene_callouts = FALSE,
  callout_n_gene_per_ctype = 5,
  callout_ctypes = NULL,
  specific_callouts = NULL,
  le_set_callouts = NULL,
  le_set_colormap = NULL,
  le_set_num_per = 5,
  show_le_legend = FALSE,
  show_xlab = TRUE,
  show_var_explained = TRUE,
  clust_method = "median",
  h_w = NULL,
  reset_other_factor_plots = FALSE,
  draw_plot = TRUE
)

Arguments

container	environment Project container that stores sub-containers for each cell type as well as results and plots from all analyses
factor_select	numeric The factor to plot
use_sig_only	logical If TRUE, includes only significant genes from jackstraw in the heatmap. If FALSE, includes all the variable genes. (default = FALSE)
nonsig_to_zero	logical If TRUE, makes the loadings of all nonsignificant genes 0 (default=FALSE)
annot	character If set to "pathways" then creates an adjacent heatmap showing which genes are in which pathways. If set to "sig_genes" then creates an adjacent heatmap showing which genes were significant from jackstraw. If set to "none" no adjacent heatmap is plotted. (default="none")
pathways	character Gene sets to plot if annot is set to "pathways" (default=NULL)
sim_de_donor_group	numeric To plot the ground truth significant genes from a simulation next to the heatmap, put the number of the donor group that corresponds to the factor being plotted (default=NULL)
sig_thresh	numeric Pvalue significance threshold to use. If use_sig_only is TRUE the threshold is used as a cutoff for genes to include. If annot is "sig_genes" this value is used in the gene significance colormap as a minimum threshold. (default=0.05)
display_genes	logical If TRUE, displays the names of gene names (default=FALSE)
gene_callouts	logical If TRUE, then adds gene callout annotations to the heatmap (default=FALSE)
callout_n_gene_per_ctype	numeric To use if gene_callouts is TRUE. Sets the number of largest magnitude significant genes from each cell type to include in gene callouts. (default=5)
callout_ctypes	character To use if gene_callouts is TRUE. Specifies which cell types to get gene callouts for. If NULL, then gets gene callouts for largest magnitude significant genes for all cell types. (default=NULL)
specific_callouts	character A vector of gene names to show callouts for (default=NULL)
le_set_callouts	character Pass a vector of gene set names to show leading edge genes for a select set of gene sets (default=NULL)
le_set_colormap	character A named vector with names as gene sets and values as colors. If NULL, then selects first n colors of Set3 color palette. (default=NULL)
le_set_num_per	numeric The number of leading edge genes to show for each gene set (default=5)
show_le_legend	logical Set to TRUE to show the color map legend for leading edge genes (default=FALSE)
show_xlab	logical If TRUE, displays the xlabel 'genes' (default=TRUE)
show_var_explained	logical If TRUE then shows an anotation with the explained variance for each cell type (default=TRUE)
clust_method	character The hclust method to use for clustering rows (default='median')
h_w	numeric Vector specifying height and width (defualt=NULL)
reset_other_factor_plots	logical Set to TRUE to set all other loadings plots to NULL. Useful if run get_all_lds_factor_plots but then only want to show one or two plots. (default=FALSE)
draw_plot	logical Set to TRUE to show the plot. Plot is stored regardless. (default=TRUE)

Value

The project container with a heatmap of loadings for one factor put in container$plots$all_lds_plots. The legend for the heatmap is put in container$plots$all_legends. Use draw(<hmap obj>,annotation_legend_list = <hmap legend obj>) to re-render the plot with legend.

Examples

test_container <- plot_loadings_annot(test_container, 1, display_genes=FALSE,
show_var_explained = TRUE)

scITD documentation built on Sept. 8, 2023, 5:11 p.m.