view_clustered_myHeatmap: Create a heatmap of analysis results.
In Benjamin-Vincent-Lab/binfotron: Binfotron Bioinformatics Analysis Tools Suite

View source: R/view_clustered_myHeatmap.R

view_clustered_myHeatmap

R Documentation

Create a heatmap of analysis results.

Description

Produces a heatmap of scaled values by row and then makes an unscaled average value column to the right of it.

Usage

view_clustered_myHeatmap(
  my_pack = NULL,
  sample_names = NULL,
  groups = NULL,
  input_df = NULL,
  base_title = NULL,
  base_file_name = NULL,
  output_dir = NULL,
  pdf_metadata = NULL,
  models_argument_list = NULL,
  imported_annotation = NULL,
  t.colors = NA,
  fileName = "cluster.cdt",
  linkage = "complete",
  distance = "pearson",
  contrast = 2,
  returnSampleClust = F,
  rowNames = NA,
  rightMar = 7,
  bottomMar = 1
)

Arguments

`my_pack`	A named list for analysis modules to send parameters to the view function in a more compact fashion: sample_names groups input_df base_file_name base_title output_dir pdf_metadata argument_list my_model annotation FDR_pValue pValue fold_change These list items may be overridden by specifying them individually. They are described in more detail below.
`sample_names`	Overrides `my_pack$sample_names`. Should be the sample names you want to have for the scaled values. They should be in the order you want them displayed. They should correspond to the rows of input_df. They should have the same order as thier groups.
`groups`	Overrides `my_pack$groups` A factored vector indicating what groups the samples belong to. Should be the same length as `sample_names`. Their order should match sample_names. They should be factored with the order matching the order in which they should be displayed. They should be formatted how you would like them displayed.
`input_df`	Overrides `my_pack$input_df` A data frame with one row per sample. Column names should correspond to the features along the y axis of the heatmap. Row order should correspond to the samples/groups should be in the order in which you would like them displayed.
`base_title`	Overrides `my_pack$base_title`. String to specify the title of the graph.
`base_file_name`	Overrides `my_pack$base_file_name`. String to specify the file name.
`output_dir`	Overrides `my_pack$output_dir`. Path to folder where the heatmap output should go. Analysis modules will determine this automatically.
`pdf_metadata`	Overrides `my_pack$pdf_metadata`. Contains data from the arguments list that should be put in the metadata of the pdf.
`models_argument_list`	Overrides `my_pack$argument_list`. Lists all of the arguments sent to the analysis module.
`imported_annotation`	Overrides `my_pack$annotation`. Annotations describe the steps taken during the analysis module and previous steps. This function will also add annotation steps to provide a full description of how the graph was generated.
`my_model`	Overrides `my_pack$my_model`. Links to all graphs are logged on a d3.js table. This variable will specify what model name was used to generate the graph on that table.
`FDR_pValue`	Overrides `my_pack$FDR_pValue`. Numeric vector with length equal to `ncol(input_df)`. If `FDR_threshhold` is not `NULL` then only features with an `FDR_pValue` less than the `FDR_threshhold` will be included.
`pValue`	Overrides `my_pack$pValue`. Numeric vector with length equal to `ncol(input_df)`. If `pValue_threshhold` is not `NULL` then only features with a `pValue` less than the `pValue_threshhold` will be included.
`fold_change`	Overrides `my_pack$fold_change`. Numeric vector with length equal to `ncol(input_df)`. If `order_by_fold_change == TRUE` then features will be ordered according to this value from high to low, where 'higher' fold change indicates group2 was higher than group1.
`order_by_fold_change`	Logical indicating whether the features should be ordered by fold change. See fold_change argument.
`base_for_log_transform`	Integer indicating the base to transform the data, which will be shown in the averages. `NULL` will results in no transformation.
`sample_panel_spacing`	`unit()` used to set the spacing between individual sample panels.
`average_panel_spacing`	`unit()` used to set the spacing between average sample panels, usually 0.
`show_sample_names`	Boolean indicating whether individual sample names should be listed along the x axis. This axis can get cluttered quickly if there are a lot of samples.
`append_classifier_to_sample_name`	Boolean indicating whether the sample group should be added to the end of the sample names. This option isn't really needed after having added facet titles.
`panel_label_switch`	'x' switches the x axis label to the bottom. `NULL` leaves the labels where they are.
`panel.strip.background`	For specifying the element_rect of the panel strip background (eg. `element_rect(fill = 'gray95')`).
`panel.strip.text.x`	For specifying the element_rect of the panel strip (eg. `element_text(size=18, angle = 0, hjust = 0.5, vjust = 0)`.
`FDR_threshhold`	If not `NULL`, features with an `FDR_pValue` above this value will not be shown.
`pValue_threshhold`	If not `NULL`, features with an `pValue` above this value will not be shown.
`display_all_names`	Boolean indicating whether the feature names should be displayed along the y axis.
`graph_rel_widths`	Numeric vector of length 2 (eg. `c(8, 2.5)`) whose ratio determines the size of the individual sample panel vs the averages.
`axis.text.x`	element_text() object to define the x axis for the sample and average plots
`axis.text.y.size`	Number indicating the size of the text for the feature names along the y axis.
`pdf_width`	Sets the width (inches) of the pdf output
`pdf_height`	Sets the height (inches) of the pdf output
`log_completion`	Binary indicting whether the graph should be added to the table of completed analyses.
`primary_title_options`	List of options to modify primary title. Title content should be set with base_title. eg: list(x=0.5,y=0.965,size=8)
`secondary_title`	List of options to modify and set secondary title. eg: list(x=1,y=0.99,size=3,label="FDR < 0.05 shown")
`plot_scale`	numeric to change how much of the page the plots take up
`legend_title`	`element_text()` object to modify the disply of the legend titles
`legend_text`	`element_text()` object to modify the disply of the legend text
`legend_key_width`	`unit()` for the width of both legends
`legend_key_height`	`unit()` for the height of both legends
`ave_legend_margin`	`margin(t, r, b, l, "pt")`
`zscore_legend_margin`	`margin(t, r, b, l, "pt")`
`ave_plot_margin`	`unit(t, r, b, l, "pt")`
`zscore_plot_margin`	`unit(t, r, b, l, "pt")`
`left_legend_pos`	string for position of zscore plot
`right_legend_pos`	string for position of average plot
`axis_title_x`	`element_text()` or `element_blank()` for axis of zscore plot
`axis_title_y`	`element_text()` or `element_blank()` for axis of zscore plot
`axis_title_x_lab`	string for axis of zscore plot, usually blank
`axis_title_y_lab`	string for aixs of zscore plot, usually blank
`zscore_colors`	vector of colors for z scale, overrides average colors if `average_scaled_vlaues == T`
`average_colors`	vecotr of colors
`average_scaled_values`	boolean indicating whether the average should be done on the raw gene expression or scaled values
`axis_text_y_remove_strings`	Strings in this character vector will be removed from the y axis text.
`axis_text_y_nchar`	Integer to specify the number of characters that will be displayed in the y axis text before clipping
`pdf_bookmark_title`	String indicating the name of the bookmark that will be used when concatentaing the pdf's into a report
`annotate_pdf`	Boolean indicating whether the pdf metadata should be chnaged. Is pdftk present?
`output_subfolder`	Path to the output sub directory folder. Multiple views of the same analysis should have unique names lest they overwrite one another. XX indicates this folder will be dropped from the TOC and bookmarks
`tile_width_factor`	Changes the size of the tiles to get rid of gaps or create them. Goes into geom_tile width parameter.
`tile_height_factor`	Changes the size of the tiles to get rid of gaps or create them. Goes into geom_tile height parameter.

Details

view_clustered_myHeatmap was created to make heatmaps of gene expression data, although it could be used to display the distribution of any numeric features. It's important to not scale data prior to heatmap to not disrupt average values.

Value

Several outputs are produced: 1) Heatmaps; 2) Heatmaps pdf; 3) An annotation file describing the steps taken on the data from start to finish; 4) A matrix file containing the sample z-scores for each sample and the average values displayed in the heatmap; 5) If log_completion == TRUE, a link is made to the pdf on the d3.js html indexing table using add_to_completed_analyses.