Dowser++, a new method of hydrating protein structures.

Area of Science:

• Computational Biology
• Structural Biology
• Biophysics

Background:

• Accurate prediction of water molecule positions in protein structures is crucial for understanding protein function and interactions.
• Existing methods like Dowser and Dowser+ have limitations in predicting hydration patterns.

Purpose of the Study:

• To introduce Dowser++, a novel computational method for predicting protein hydration structures.
• To evaluate the performance of Dowser++ against experimental data and compare it with previous methods.

Main Methods:

• Dowser++ is a semi-empirical modification of the Dowser program, incorporating AutoDock Vina and WaterDock protocols.
• Predictions were validated using experimental data from 14 oligopeptide-binding protein (OppA) crystal structures, cytochrome c oxidase channels, and PSII data.
• A quantitative model was developed to assess the quality of predicted water molecule placements, accounting for potential crystallographic false negatives.

Main Results:

• Dowser++ demonstrated significantly improved predictive power compared to the original Dowser and Dowser+.
• The method accurately predicted water molecule positions in diverse protein systems, including those with extensive internal water networks.
• The developed quality estimation model provides a robust assessment of prediction accuracy.

Conclusions:

• Dowser++ represents a substantial advancement in computational protein hydration prediction.
• The enhanced accuracy of Dowser++ facilitates more reliable structural and functional analyses of proteins.
• This method offers a valuable tool for researchers in structural biology and computational biophysics.