MAB, a generally applicable molecular force field for structure modelling in medicinal chemistry

Area of Science:

• Computational Chemistry
• Molecular Modeling
• Structural Biology

Background:

• Existing molecular force fields often require complex parameterization and partial atomic charge assignments.
• The accurate prediction of molecular structures is crucial for understanding biological processes and designing new materials.

Purpose of the Study:

• To present a new, generally applicable molecular force field with simplified parameterization.
• To introduce a novel approach for handling hydrogen bonding and electrostatic interactions.
• To evaluate the accuracy and performance of the new force field for diverse molecular structures.

Main Methods:

• Developed a new mathematical formulation and parametrization scheme for a molecular force field.
• Employed a consistent united-atom approximation, omitting explicit hydrogen atoms.
• Replaced electrostatic terms with geometrical hydrogen-bonding terms and utilized dynamic network analyses for hydrogen-bond pattern computation.
• Tested the force field on 1589 structures from the Cambridge Structural Database and qualitatively on proteins and DNA oligomers.

Main Results:

• Achieved average agreements of 2 pm for bond lengths and 2 degrees for valence angles.
• Demonstrated root-mean-square deviations of 10 to 20 pm for atomic positions on tested structures.
• Showcased satisfactory performance in predicting large-scale structural properties of proteins and DNA oligomers.

Conclusions:

• The new molecular force field offers a simplified yet accurate approach to molecular modeling.
• The method's ability to dynamically compute hydrogen-bond patterns and its finite-range design enhance its applicability.
• This force field shows significant potential for structural analysis across various molecular systems, including biomacromolecules.