The MS2DB ++ webserver: disulfide bond determination through evidence combination.

Area of Science:

• Computational biology
• Structural bioinformatics
• Proteomics

Background:

• Disulfide bonds are crucial for protein structure and function.
• Accurate determination of disulfide connectivity is essential for understanding protein folding and stability.
• Existing computational methods have limitations in predicting disulfide topology for complex proteins.

Purpose of the Study:

• To develop an integrated web server, MS2DB++, for enhanced prediction of protein disulfide connectivity.
• To combine evidence from mass spectrometry and sequence-based methods with user-defined external tools.
• To improve the reliability of disulfide bond prediction, especially for challenging molecular structures.

Main Methods:

• MS2DB++ web server integrates mass spectrometry-based and two protein sequence-based predictive methods.
• Allows incorporation of up to two external prediction methods.
• Employs Dempster-Shafer theory with four evidence combination formulations for robust analysis.

Main Results:

• MS2DB++ provides a unified platform for disulfide connectivity prediction.
• The integrated approach enhances prediction accuracy compared to single methods.
• Demonstrates utility in complex scenarios where individual methods show inconsistencies.

Conclusions:

• MS2DB++ offers a powerful and flexible tool for predicting protein disulfide connectivity.
• The Dempster-Shafer theory-based evidence combination improves prediction robustness.
• This approach is particularly valuable for structural and functional proteomics research.