αCharges: partial atomic charges for AlphaFold structures in high quality

Area of Science:

• Structural biology
• Computational chemistry
• Bioinformatics

Background:

• AlphaFold2 has generated millions of protein structure predictions, available in AlphaFoldDB.
• These structures lack detailed functional annotations, such as partial atomic charges, which are crucial for understanding chemical reactivity.
• Existing methods for charge calculation are not optimized for the scale of AlphaFoldDB data.

Purpose of the Study:

• To introduce αCharges, a web application for calculating partial atomic charges for AlphaFoldDB protein structures.
• To provide essential chemical behavior data for large-scale predicted protein structures.
• To facilitate the use of predicted protein structures in downstream chemical and biological analyses.

Main Methods:

• Developed the αCharges web application.
• Employed the SQE+qp empirical method for charge calculation.
• Utilized quantum mechanics (B3LYP/6-31G*/NPA) derived charges on PROPKA3 protonated structures for parameterization.
• Integrated with the Mol* viewer for visualization and common data formats for download.

Main Results:

• Successfully calculated partial atomic charges for millions of protein structures from AlphaFoldDB.
• Provided a user-friendly web interface for accessing and visualizing charge data.
• Enabled download of charge data in standard formats.

Conclusions:

• αCharges significantly enhances the functional annotation of predicted protein structures.
• The tool democratizes access to crucial chemical information for large protein structure datasets.
• αCharges will aid research in protein function, drug design, and chemical biology.