Flexible fitting in 3D-EM with incomplete data on superfamily variability

Area of Science:

• Structural Biology
• Biophysics
• Computational Biology

Background:

• Flexible fitting in 3D electron microscopy (3D-EM) is crucial for understanding protein dynamics.
• Previous S-flexfit methods relied on Singular Value Decomposition (SVD), limited by sequence alignment gaps.

Purpose of the Study:

• To enhance the S-flexfit method for improved flexible fitting in 3D-EM.
• To incorporate Incremental Singular Value Decomposition (ISVD) to handle missing data in structural alignments.

Main Methods:

• Developed an improved S-flexfit incorporating ISVD, an extension of SVD designed for datasets with missing values.
• Utilized evolutionary information from protein domain databases (e.g., CATH) for structural alignment of protein superfamily members.
• Applied the enhanced S-flexfit method to both simulated and experimental 3D-EM maps.

Main Results:

• ISVD effectively manages gaps in structural alignments, allowing consideration of more amino acids.
• The enhanced S-flexfit demonstrates improved performance in flexible fitting compared to the previous SVD-based approach.
• Successful application to GroEL and Poliovirus 135S cell entry intermediate 3D-EM maps.

Conclusions:

• The ISVD-based S-flexfit offers a more robust and broadly applicable method for flexible fitting in 3D-EM.
• This advancement facilitates more accurate structural modeling, especially with diverse and gapped sequence data.
• The improved method provides better models for interpreting 3D-EM data and understanding protein structures.