A comprehensive package for accurate protein crystal structure refinement

The Advantage of X-ray Crystal Structure Refinement

The prevailing geometric restraints employed in protein crystallography apply experimental bond length and angle terms as well as other restraint terms that have been subsequently added. However, some potential issues arise when refined structures are used in downstream computational modeling.

Two key characteristics of protein crystal structures that could affect the accuracy of subsequent structure-based modeling are:

  • High-energy contacts interfere with computational chemistry calculations, and are often removed by the application of restrained energy minimization to the crystal structure; the danger with this procedure is the introduction of changes in the structure not supported by the X-ray data.
  • Most protein crystal structures at typical resolutions do not include hydrogens in the model, which must be added after the end of refinement for many molecular mechanics calculations.

Traditionally, attempts to remediate the aforementioned issues are done after refinement, which shifts the control of structural results away from the scientists who are most familiar with the interpretation of diffraction experiments. PrimeX directly addresses these concerns by restraining protein geometry to OPLS-AA (one of the most accurate and widely-deployed force fields for studying protein/ligand systems) during X-ray refinement, and by adding hydrogens during refinement and fully accounting for their existence in all energy computations.

Furthermore, just as the inclusion of hydrogen atoms provides important information for structure validation of refinement results, PrimeX also features improved accounting of non-bonded interactions during refinement, which are central to understanding ligand binding. Thus, PrimeX provides a complete environment that facilitates refinement and produces accurate structures more compatible with computational chemistry applications than those produced by other protein refinement programs.

Loop building and refinement:
PrimeX builds loops up to 40-residues in length, using technologies in the well-validated Prime protein modeling program and guided by electron density fit.

Ligand placement:
PrimeX places ligands and other small molecules into electron density using technologies in the Glide docking program, which has demonstrated superior accuracy in ligand-receptor docking.

Accurate all-atom force field:
PrimeX utilizes the OPLS-AA force field with state-of-the-art computational technologies to refine protein structures that are immediately ready for all computational simulations.

Advanced refinement techniques:
PrimeX provides simulated annealing for reciprocal space refinement.

Choice of minimizers:
PrimeX offers conjugate gradient, truncated Newton, and quasi-Newton (LBFGS) to optimize performance and accuracy.

Automatic parameter generation:
PrimeX generates parameters for ligands and other small molecules, as well as modified residues, automatically without requiring user intervention.

Treatment of hydrogens:
PrimeX automatically adds hydrogens, which are included during refinement according to physical chemistry as prescribed by the OPLS-AA force field.

Easy to use:
PrimeX's intuitive user interface is integrated into Maestro with step-by-step organization of refinement statistics in the Project Table and convenient analysis of protein structure geometry through interactive tables and plots.

Advanced calculational controls:
PrimeX allows command-line input as well as scripting with Python for added control and customizable operations.

Citations and Acknowledgements

Schrödinger Release 2023-4: PrimeX, Schrödinger, LLC, New York, NY, 2023.

ö Bell, J.A., Cao, Y., Gunn, J.R., Day, T., Gallicchio, E., Zhou, Z., Levy, R. and Farid, R., "PrimeX and the Schrödinger Computational Chemistry Suite of Programs," International Tables for Crystallography, Volume F, Crystallography of Biological Macromolecules, 2012, 18, 534-538

· "Investigating Protein–Peptide Interactions Using the Schrödinger Computational Suite"

Bhachoo, J.; Beuming, T., Methods Mol Biol., 2017, 1561, 235-254

"A Structure-Based Model for Predicting Serum Albumin Binding"

Lexa, K. W.; Dolghih, E.; Jacobson, M. P., PLoS ONE, 2014, 9(4), e93323

"Structure of the Arabidopsis thaliana TOP2 oligopeptidase"

Wang, R.; Rajagopalan, K.; Sadre-Bazzaz, K.; Moreau, M.; Klessig, D. F.; Tong, L., Acta. Crystallogr. F Struct. Biol. Commun., 2014, 70(Pt 5), 555-559

· "Allosteric Inhibition of the NS2B-NS3 Protease from Dengue Virus"

Yildiz, M.; Ghosh, S.; Bell, J. A.; Sherman, W.; Hardy, J. A., ACS Chem. Biol., 2013, 8(12), 2744-2752

· "PrimeX and the Schrödinger computational chemistry suite of programs"

Bell, J. A.; Cao, Y.; Gunn, J. R.; Day, T.; Gallicchio, E.; Zhou, Z.; Levy, R.; Farid, R., International Tables for Crystallography, Volume F: Crystallography of biological macromolecules, 2012, 18, 534-538

· "Significant reduction in errors associated with non-bonded contacts in protein crystal structures: Automated all-atom refinement with PrimeX"

Bell, J. A.; Ho, K. L.; Farid, R., Acta. Crystallogr. D Biol. Crystallogr., 2012, 68(Pt 8), 935-952

"The crystal structure of DehI reveals a new α-haloacid dehalogenase fold and active-site mechanism"

Schmidberger, J. W.; Wilce, J. A.; Weightman, A. J.; Whisstock, J. C.; Wilce, M. C., J. Mol. Biol., 2008, 378, 284-294
Back To Top