FOCAL

FOCAL implements a grid-based clustering algorithm to detect and refine clusters in single-molecule localization microscopy (SMLM) datasets for accurate identification of sub-cellular assemblies.

Key Features:

• Grid-Based Clustering Algorithm: Employs a grid-based approach for localization clustering that processes data with linear computational scaling O(n) in the number of localizations, contrasted with DBSCAN's O(n log(n)) scaling.
• Parameter Efficiency: Uses a single set parameter to reduce subjectivity and complexity in parameter selection compared with multi-parameter algorithms such as DBSCAN.
• Artifact Management: Explicitly accounts for common SMLM reconstruction artifacts to reduce detection of false or pseudo clusters.
• Out-of-Focus Cluster Filtering: Implements filtering of out-of-focus clusters to improve the reliability of cluster identification in complex SMLM images.

Scientific Applications:

• dSTORM — RNAP II clusters: Applied to dSTORM datasets to analyze eukaryotic RNA polymerase II (RNAP II) cluster organization.
• PALM — H-NS clustering: Applied to PALM datasets to assess clustering of the bacterial protein H-NS.

Methodology:

Implements a grid-based clustering algorithm and was evaluated via comparative analysis with DBSCAN on experimental and simulated SMLM datasets, demonstrating O(n) scaling versus DBSCAN's O(n log(n)).