Point cloud deformation modeling for particle selection following cryo-EM 2D classification

Key Points

• The aim is to improve the quality of particle datasets by identifying and removing deformed or misclassified particles.
• Develop a model for point cloud deformation analysis.
• Implement a data-filtering step after standard 2D classification procedures.
• Evaluate the model's impact on the reliability of particle datasets.
• Successfully identifies deformed or misclassified particles, improving dataset quality.
• Enhances the reliability of subsequent cryo-EM analysis and interpretations.