Leopard-EM

Leopard-EM implements modular two-dimensional template matching (2DTM) to locate and orient molecular complexes in cryo-EM images with high precision for in situ structural biology.

Key Features:

• Modular Architecture: The software provides an extensible Python-based framework that supports rapid development of novel 2DTM methodological approaches and customization of in situ cryo-EM data processing workflows.
• Automated Pixel-Size Refinement: The package implements a procedure to refine pixel size, as 2DTM sensitivity is affected by pixel-size errors of approximately 0.2%.
• Constrained Search Protocol: The architecture supports constrained searches using initial locations and orientations from related structures, demonstrated by an approximately eightfold improvement in small ribosomal subunit detection when using large ribosomal subunit positions.

Scientific Applications:

• Single-Molecule Annotation in Cellular Cryo-EM: 2DTM uses high-resolution structural models as computational probes to locate and orient molecular complexes within cellular cryo-EM images at single-molecule resolution.
• Ribosome State Analysis: The constrained search strategy captures distributions of ribosome rotation states within living cells at single-molecule resolution.
• In Situ Structural Biology Platform: The software serves as a development platform for novel in situ cryo-EM data-processing workflows and methodological innovations.

Methodology:

Leopard-EM performs 2D template matching using high-resolution structural models as computational probes. The implementation includes an automated pixel-size refinement procedure and supports constrained search protocols that use initial locations and orientations from one structure to guide detection of another.