DEnote

Interactive cell labeling for scRNA-seq data. Built to make a key stage in scRNA-seq analysis accessible to non-bioinformaticians.

Disclaimer: DEnote was initially created while benchmarking / experimenting with Posit AI interfacing with Claude Sonnet 4.6

Functionality

• Supports Seurat (.rds) or 10x (.h5) files
• Performs Leiden or Louvain clustering
• Explore 2D/3D UMAP, tSNE, and PCA reductions
• Visualize gene expression across reductions, or violin plots
• De novo marker analysis via FindAllMarkers()
• Score known markers as modules via logistic regression
• Calculate inter-cluster similarity via Bhattacharyya coefficient (BC)
• Manually or automatically label clusters based on visual and statistical criteria
• Export easy to read html reports and .csv of metadata and differential expression results

User Start Guide

1. Clone github repository

git clone https://github.com/Tripfantasy/DEnote.git

2. Download dependencies in R (R/Rstudio required) app built using R v4.5.3

packages <- c("shiny", "bslib", "Seurat", "plotly", "DT", "ggplot2", 
              "ggrepel", "dplyr", "shinycssloaders", "shinyjs", 
              "rmarkdown", "scales", "RColorBrewer", "tidyr", 
              "tibble", "thematic","bsicons")

install_if_missing <- function(p) {
  if (!requireNamespace(p, quietly = TRUE)) install.packages(p)
}

lapply(packages, install_if_missing)

3. Run App

shiny::runApp("path/to/app.R")

Previews