Repository files navigation

visxhclust is a package that consists mainly of a Shiny application forvisual exploration ofhierarchicalclustering. Itimplements hierarchical clustering for numeric data and includes variousplots and interactions to facilitate iterative workflows. Featuresinclude:

• Selection of variables supported by a correlation-driven t-test
• Visual inspection of MDS and PCA projections
• Visualisation of results with dendrogram, heatmap and boxplots
• Validation by significance testing, internal validation scores andindices and computation of Gap statistic
• Comparison of multiple clustering results with Sankey plots
• Saving and loading of settings and results

The app includes multiple help points in the interface, and the packageadditionally exports various functions to help with documenting andreproducing a clustering workflow in R Markdown – checkvignette("clusterworkflow") for an example.

The package can be installed from github using:

remotes::install_github("rhenkin/visxhclust")

Most packages are from CRAN. However, the heatmap drawing package ispart ofBioconductor and may require aseparate installation:

install.packages("BiocManager")BiocManager::install("ComplexHeatmap")

To use your data with the tool, you can save a data frame or tibble inan RDS file, or use comma or tab-delimited files, with .csv, .tsv or.txt extensions. The clustering in the tool is done on numeric values;columns containing text will be listed under possible heatmapannotations. If an ID column exists, it will be used as an internalidentifier for rows, as the app support removing rows through the tableview.

Clustering requires complete datasets with no empty values, NULLs orNAs. If any column contains missing values, it will be set aside to beused as a heatmap annotation. Badly formatted data will also lead tounexpected results in the tool. Imputation packages can be used to fillmissing data and faulty rows should be removed before loading the fileinto the tool. The tool is not intended to help with this stage of dataprocessing.

To run the app once the package is installed:

library(visxhclust)run_app()