OPUS-SSRI

OPUS-SSRI performs Maximum A Posteriori refinement of cryo-electron microscopy (cryo-EM) single-particle 3D reconstructions by imposing spatially varying sparseness and smoothness priors to enhance electron density map quality and resolution for structural biology.

Key Features:

• Incorporation of Spatially Varying Priors: OPUS-SSRI integrates spatially varying sparseness and smoothness priors into the refinement to regularize electron density maps.
• Real Space Penalty Function: The method employs a real space penalty function to enforce the priors during refinement, reducing noise and artifacts.
• Local Kernel Regression Back-Projection: Back-projection is redefined as a local kernel regression to improve alignment of projections with the 3D model.
• First-Order Optimization Method: A first-order optimization method is used to solve the refinement problem efficiently.

Scientific Applications:

• Cryo-EM single-particle analysis: OPUS-SSRI improves electron density map resolution within cryo-EM single-particle reconstruction workflows.
• Atomic-level structural determination: Enhanced map quality and resolution support atomic-level interpretation of macromolecular complexes and molecular interactions.

Methodology:

OPUS-SSRI performs Maximum A Posteriori (MAP) refinement of 3D reconstructions using spatially varying sparseness and smoothness priors enforced by a real-space penalty function, implements a local kernel regression back-projection, and solves the optimization with a first-order method; comparative tests on seven cryo-EM datasets reported an average resolution improvement of 0.64 Å and a maximum improvement of 1.25 Å.