High-resolution structure prediction of β-barrel membrane proteins

Area of Science:

• Structural Biology
• Biochemistry
• Computational Biology

Background:

• Beta-barrel membrane proteins (BMPs) are crucial but their structures are poorly understood.
• Existing methods struggle with predicting BMP structures, especially for novel or underrepresented protein families.

Purpose of the Study:

• To develop an accurate computational method for predicting the 3D structures of BMPs.
• To improve the prediction of transmembrane domains, loops, and overall barrel structures.

Main Methods:

• A novel computational approach to predict strand registers and construct transmembrane domains.
• Application to diverse BMPs, including those with limited sequence data and novel folds.
• Validation against experimentally resolved structures (X-ray crystallography and NMR).

Main Results:

• Achieved an average main-chain RMSD of 3.48 Å for predicted transmembrane domains, a significant improvement over previous methods.
• Demonstrated high accuracy for BMPs with NMR structures, comparable to experimental variability.
• Successfully modeled extended beta-barrels and loops, increasing prediction coverage by 20%.

Conclusions:

• The developed method provides accurate 3D structure predictions for a wide range of BMPs.
• This approach enhances structural coverage and can be applied genome-wide.
• Enables deeper understanding of BMP function and facilitates drug discovery efforts.