Skip to contents

HPAclusteR

HPAclusteR is an R package developed by the Human Protein Atlas to perform gene clustering from transcriptomics data. The package provides a complete pipeline for clustering genes, visualizing clustering results, and annotating gene clusters with functional databases such as Gene Ontology and others. It is designed to facilitate the analysis of transcriptomics data and help researchers identify biologically meaningful gene clusters.

Note: This package is built to work with AnnDatR objects, which are designed to handle transcriptomics data efficiently. The package that conceptualizes this data structure can be found here.

Features

Gene Clustering Pipeline:

  • Perform PCA for dimensionality reduction.
  • Calculate distances between genes.
  • Construct Shared Nearest Neighbor (SNN) graphs.
  • Perform consensus clustering to identify robust gene clusters.
  • Generate UMAP embeddings for visualization.
  • Create UMAP cluster hulls to highlight cluster boundaries.

Annotate Gene Clusters with Functional Databases:

  • Gene Ontology (GO)
  • KEGG Pathways
  • Humnan Protein Atlas (HPA)
    • Tissue Expression
    • Single Cell Type Expression
    • Blood Expression
    • Brain Expression
    • Subcellular Location
    • Secretome Location
    • Protein Class
  • Reactome Pathways
  • PanglaoDB Cell Markers
  • Trrusted Cell Markers

Ready-to-use Visualization Functions:

  • Plot UMAP embeddings with clusters and/or hulls.
  • Plot gene expression per cluster across samples.
  • Plot annotation results with treemaps and buble heatmaps.
  • Plot cluster comparison results with network graphs.

Installation

You can install the development version of HPAclusteR from GitHub:

# Install devtools if not already installed
install.packages("devtools")

# Install HPAclusteR
devtools::install_github("buenoalvezm/HPAclusteR")

Usage

Start right away with the hc_auto_cluster() function, which performs the complete gene clustering pipeline on an AnnDatR object.

library(HPAclusteR)

# Example input: AnnDatR object with transcriptomics data
adata_res <- hc_auto_cluster(example_adata, cluster_resolution = 10)

Issues and Support

If you encounter any bugs or you want to recommend new features and changes to existing ones, please open a new issue on our GitHub repository.

Contact

For any questions or further information, please contact us at k.antono@outlook.com.