Density Peaks Advanced clustering in R

The DPAclustR package is a wrapper to the Python library DPA implementing the Density Peaks Advanced (DPA) clustering algorithm as introduced in the paper "Automatic topography of high-dimensional data sets by non-parametric Density Peak clustering", published on M. d'Errico, E. Facco, A. Laio, A. Rodriguez, Information Sciences, Volume 560, June 2021, 476-492 (also available on arXiv).

The package offers the following features:

• Advanced version of the DP clustering algorithm, published in the Clustering by fast search and find of density peaks paper, which includes an automatic search of cluster centers and assessment of statistical significance of the clusters.

Source files

The source R codes are inside the R folder:

.
|-- ...
|-- R/
|   |-- Plotting.R                   # Visualization tools:
|   |                                # dendrogram and network visualizations.
|   |
|   |-- AnalysisTools.R              # Normalized Mutual Information score,
|   |                                # grid search for choosing of
|   |                                # the external parameters
|   |
|   |-- DPAclustering.R              # Function running the Density Peak Advanced clustering.
|
|
|-- ...

Getting started

The source code of DPAclustR is on github DPAclustR repository.

We suggest you to use renv as project environment for R. The renv.lock file can be used to restore a working status of the DPAclustR library by running the command renv::restore(). Alternatively, a new project environment can be initialized with the command renv::init(). See Installation section for more details.

The DPA Python package is required to use the DPAclustR::runDPAclustering function. Please see intruction on github DPA repository for installing it. For more details on the runDPAclustering and its usage see Quickstart section below.

Installation

Assuming you already have R or RStudio installed on your machine, run the following commands to install the SingleCellAnalysis package from github:

renv::init() # suggested
install.packages("devtools")
devtools::install_github("mariaderrico/DPAclustR")

For development, you can clone the DPAclustR source code:

git clone https://github.com/mariaderrico/DPAclustR.git

The renv.lock file in the package can be used to restore a working status of the DPAclustR library by running the following commands:

setwd("yourlocalpath/DPAclustR")
renv::restore()
library(DPAclustR)

Quickstart

A use-case example of the analysis workflow is provided as jupyter notebook in Analysis_example.ipynb. The same analysis workflow is provided as R script in Analysis_example.R that can be easily run within RStudio.

To run the clustering analysis using the DPA method, as described under section Analysis tools of the use-case example, the DPA Python package has to be loaded. Assuming the DPA package has been installed following the instructions available in the github DPA repository, using virtualenv, the following commands have to be run:

library(reticulate)
use_virtualenv("path_toEnv/venvdpa/", required=TRUE)
setwd("path_toDPA/DPA")
DPA <- import_from_path("DPA", path="src/Pipeline/")