Skip to content

Latest commit

 

History

History
81 lines (68 loc) · 3.75 KB

notes.md

File metadata and controls

81 lines (68 loc) · 3.75 KB

Notes and changes

  • Disable zoom in scatterD3.
    • In d3-4.13.0.min.js, change the wheel.zoom to null.
  • Change color scale in scatterD3 and export functions to YlGnBu.
    • Required adding d3-scale-chromatic.v.0.3.min.js to scatterD3 and change object name inside.
    • Eventually consider making this a variable to choose by user, e.g. blue vs red scale.
  • Make own box function so I don't have to repeat parameters all the time?
    • Same for action button.
    • Also modalDialog.
  • Remove as.vector() for gene expression, not necessary.
  • Check/update export scripts.
  • Move to z-score instead of log-counts?
    • Not possible because occupy much more data.
  • Use formattable.
  • Make only RDS files selectable.
  • Change text in some panels.
  • Make input in gene expression panel case-independent.
  • Make refresh shortcut local (instead of global).
  • Make sample names be displayed instead of the long names (top expressed genes and marker genes).
  • Border around plots.
  • Show names in sample-by-cluster and cluster-by-sample plots and cell cycle in respective plots.
  • Allow user to choose plotting order.
  • Reorder some user inputs to save space.
  • Show cluster tree if available.
  • Hide some columns from tables in enriched pathways to have a cleaner view. They can still be shown manually.
  • Use a box layout across the app.
  • Use more attractive colors (from <flatuicolors.com>).
  • Add info popup boxes with additional explanation for each/most panel(s).
  • Reduce empty space (padding around boxes) across the app.
  • Remove Z-score from marker gene tables since combined score should be more informative.
  • Add color bar representation through formattable in marker gene tables.
  • Remove legend in gene expression and gene set expression scatter plots because scale will still be visible in box plots.
  • Add checks for presence of tables of top expressed genes and marker genes.
  • Hide gene column in enriched pathway tab by default to save space.
  • Update R version for Windows to 3.5.1.
  • Add export buttons for dimensional reductions using ggplot instead of the scatterD3 function which crashes with too many cells.
  • Manually modify code for scatterD3 legend to prevent sorting. A more consistent solution should be found however.
  • If marker genes haven't already been checked in the pipeline, this will be done in the browser for murine and human samples.

How to uninstall all non-base R packages

Taken from: https://www.r-bloggers.com/how-to-remove-all-user-installed-packages-in-r/

# create a list of all installed packages
ip <- as.data.frame(installed.packages())

# we don't want to remove base or recommended packages either\
ip <- ip[!(ip[,"Priority"] %in% c("base", "recommended")),]

# determine the library where the packages are installed
path.lib <- unique(ip$LibPath)

# create a vector with all the names of the packages you want to remove
pkgs.to.remove <- ip[,1]

# remove the packages
sapply(pkgs.to.remove, remove.packages, lib = path.lib)

Then, re-install the libraries required by Cerebro. The maintenance of portable R installations is tricky and I'm probably not doing this in a good way.

On macOS, in my experience, it works best to run a "real" R installation, assign the library path of the portable installation and then install the cerebroApp package there with BiocManager::install. Trying to install packages from within the portable R usually fails during compilation.

assign(".lib.loc", "...", envir = environment(.libPaths))
install.packages("BiocManager")
install.packages("remotes")
BiocManager::install("romanhaa/cerebroApp")

On Windows, it usually works fine to run the portable R installation and install cerebroApp from there.

install.packages("BiocManager")
install.packages("remotes")
BiocManager::install("romanhaa/cerebroApp")