HDCluster Documentation

📘 Table of Contents

• 🧩 Overview
• ⚙️ Stand Alone Windows Application Installation
• 📂 Supported Data Formats
• 🚀 Using HDCluster
• 🎬 Demo Videos

🧩 Overview

HDCluster is available as a preprint on bioRxiv (DOI: https://doi.org/10.1101/2025.10.23.684134).

HDCluster is a simple, yet powerful, one-parameter spatial clustering method. It is designed to be scalable, robust to noise, and exceptionally fast, making it suitable for handling very large datasets, such as those generated by DNA-PAINT Single Molecule Localization Microscopy (SMLM).

HDCluster can be used as a standalone tool with a denoising option, or as a module within a larger computational pipeline to retrieve clustered localizations. Its versatility allows it to be applied to both biological and non-biological data.

One of the key advantages of HDCluster is its minimal calibration requirement. It relies on a single parameter that can be set based on the physical characteristics of the clusters, such as their size. This parameter can be determined by analyzing the scale of clustering with a global Ripley's H-function or by measuring the spread of a known cluster within the data.

HDCluster's strength lies in its ability to identify clusters of various shapes and densities while automatically filtering out noisy localizations. By combining density-based clustering with graph-based techniques, HDCluster provides robust and accurate cluster identification in SMLM datasets, which is crucial for precise emitter reconstruction and other super-resolution microscopy applications.

HDCluster is available in both Python and MATLAB.

⚙️ Standalone Windows Application Installation

A compiled version of HDCluster is available for Windows. The installation process is straightforward.

1. Locate the HDCluster_App_Installer file and double-click it to begin the installation.
2. Follow the setup wizard to install HDCluster on your computer.

HDCluster requires the MATLAB Runtime (MCR), which will be installed automatically if it is not already present on your system. The MCR is a freely available set of shared libraries required to run the software. Please note that administrator privileges are required to install the MCR.

Windows Installation Guide

You can find the Windows installation guide here.

📂 Supported Data Formats

HDCluster supports a variety of file formats, including:

• .txt
• .hdf5
• .csv
• .xyz
• .3dlp
• .ascii

Regardless of the file format, the file must contain a header with acquisition information for the localizations. For 2D data, the file must include X and Y columns. For 3D data, X, Y, and Z columns are required. Please ensure that the column labels do not contain prefixes or postfixes (e.g., use X instead of X (nm)).

Below are examples of sample files in both 2D and 3D formats. Each example illustrates the file header, which contains metadata and parameters, followed by a preview of the first few rows of localization data.

• Sample .txt file data

• Sample .csv file data

🚀 Using HDCluster

You can find a guide for using HDCluster here.

📺 Watch Demo Videos Here

You can find all demo videos in the Demo Videos section.

Watch Supplementary Videos Here

You can find all demo videos in the Supplementary Videos section.

Original Developmental Repository and Developer

Link to repository