Predicting the solvent accessibility of transmembrane residues from protein sequence

Area of Science:

• Biochemistry
• Structural Biology
• Computational Biology

Background:

• Membrane proteins are crucial biological components, but their experimental structure determination is challenging.
• Understanding the tertiary structure and residue accessibility is vital for elucidating protein function.
• Transmembrane residues interact with lipid environments, influencing protein stability and function.

Purpose of the Study:

• To develop a novel computational method for predicting solvent accessible surface areas (SASA) of transmembrane residues.
• To enable the identification of residues exposed to the lipid environment versus those buried within the protein.
• To provide a tool for theoretical modeling of membrane protein structures, especially transmembrane domains.

Main Methods:

• Development of a predictive model for solvent accessible surface areas.
• Application of the method to both transmembrane alpha-helix and beta-barrel residues.
• Extensive validation using different SASA definitions and parameter sets.

Main Results:

• Achieved correlation coefficients of approximately 0.65 between predicted and observed SASA for both alpha-helix and beta-barrel residues.
• Demonstrated the ability to differentiate between lipid-exposed and buried residues using predicted SASA values.
• Validated the robustness of the prediction method through comprehensive testing.

Conclusions:

• The proposed method offers a valuable computational approach for predicting transmembrane residue accessibility.
• This tool can significantly aid in the theoretical modeling of membrane protein tertiary structures and domain assembly.
• The method facilitates the annotation of membrane proteins within proteomes, offering insights into structural and functional properties.