High-throughput quantum-mechanics/molecular-mechanics (ONIOM) macromolecular crystallographic refinement with PHENIX/DivCon: the impact of mixed Hamiltonian methods on ligand and protein structure.

Conventional macromolecular crystallographic refinement relies on often dubious stereochemical restraints, the preparation of which often requires human validation for unusual species, and on rudimentary energy functionals that are devoid of nonbonding effects owing to electrostatics, polarization, charge transfer or even hydrogen bonding. While this approach has served the crystallographic community for decades, as structure-based drug design/discovery (SBDD) has grown in prominence it has become clear that these conventional methods are less rigorous than they need to be in order to produce properly predictive protein-ligand models, and that the human intervention that is required to successfully treat ligands and other unusual chemistries found in SBDD often precludes high-throughput, automated refinement. Recently, plugins to the Python-based Hierarchical ENvironment for Integrated Xtallography (PHENIX) crystallographic platform have been developed to augment conventional methods with the in situ use of quantum mechanics (QM) applied to ligand(s) along with the surrounding active site(s) at each step of refinement [Borbulevych et al. (2014), Acta Cryst D70, 1233-1247]. This method (Region-QM) significantly increases the accuracy of the X-ray refinement process, and this approach is now used, coupled with experimental density, to accurately determine protonation states, binding modes, ring-flip states, water positions and so on. In the present work, this approach is expanded to include a more rigorous treatment of the entire structure, including the ligand(s), the associated active site(s) and the entire protein, using a fully automated, mixed quantum-mechanics/molecular-mechanics (QM/MM) Hamiltonian recently implemented in the DivCon package. This approach was validated through the automatic treatment of a population of 80 protein-ligand structures chosen from the Astex Diverse Set. Across the entire population, this method results in an average 3.5-fold reduction in ligand strain and a 4.5-fold improvement in MolProbity clashscore, as well as improvements in Ramachandran and rotamer outlier analyses. Overall, these results demonstrate that the use of a structure-wide QM/MM Hamiltonian exhibits improvements in the local structural chemistry of the ligand similar to Region-QM refinement but with significant improvements in the overall structure beyond the active site.

Assessment of detailed conformations suggests strategies for improving cryoEM models: Helix at lower resolution, ensembles, pre-refinement fixups, and validation at multi-residue length scale.

We find that the overall quite good methods used in the CryoEM Model Challenge could still benefit greatly from several strategies for improving local conformations. Our assessments primarily use validation criteria from the MolProbity web service. Those criteria include MolProbity's all-atom contact analysis, updated versions of standard conformational validations for protein and RNA, plus two recent additions: first, flags for cis-nonPro and twisted peptides, and second, the CaBLAM system for diagnosing secondary structure, validating Cα backbone, and validating adjacent peptide CO orientations in the context of the Cα trace. In general, automated ab initio building of starting models is quite good at backbone connectivity but often fails at local conformation or sequence register, especially at poorer than 3.5 Šresolution. However, we show that even if criteria (such as Ramachandran or rotamer) are explicitly restrained to improve refinement behavior and overall validation scores, automated optimization of a deposited structure seldom corrects specific misfittings that start in the wrong local minimum, but just hides them. Therefore, local problems should be identified, and as many as possible corrected, before starting refinement. Secondary structures are confusing at 3-4 Šbut can be better recognized at 6-8 Å. In future model challenges, specific steps being tested (such as segmentation) and the required documentation (such as PDB code of starting model) should each be explicitly defined, so competing methods on a given task can be meaningfully compared. Individual local examples are presented here, to understand what local mistakes and corrections look like in 3D, how they probably arise, and what possible improvements to methodology might help avoid them. At these resolutions, both structural biologists and end-users need meaningful estimates of local uncertainty, perhaps through explicit ensembles. Fitting problems can best be diagnosed by validation that spans multiple residues; CaBLAM is such a multi-residue tool, and its effectiveness is demonstrated.

Morphing and docking visualisation of biomolecular structures using Multi-Dimensional Scaling.

Protein structures are often solved at atomic resolution in two states defining a functional movement but intervening conformations are usually unknown. Morphing methods generate intervening conformations between two known structures. When viewed as an animation using molecular graphics, a smooth, direct morph enables the eye to track changes in structure that might be otherwise missed. We present a morphing method that aims to linearly interpolate interatomic distances and which uses SMACOF (Scaling by MAjorisation of COmplicated Function) and multigrid techniques with a cut-off distance based weighting that optimizes the MolProbity score of intervening structures. The all-atom morphs are smooth, move directly between the two structures, and are shown, in general, to pass closer to a set of known intermediates than those generated using other methods. The techniques are also used for docking by putting the unbound structures in a "near-approach pose" and then morphing to the bound complex. The resulting GPU-accelerated tools are available on a webserver, Morphit_Pro, at http://morphit-pro.cmp.uea.ac.uk/ and more than 5000 domains movements available at the DynDom website can now be viewed as morphs http://morphit-pro.cmp.uea.ac.uk/dyndom/.

Model validation: local diagnosis, correction and when to quit.

Traditionally, validation was considered to be a final gatekeeping function, but refinement is smoother and results are better if model validation actively guides corrections throughout structure solution. This shifts emphasis from global to local measures: primarily geometry, conformations and sterics. A fit into the wrong local minimum conformation usually produces outliers in multiple measures. Moving to the right local minimum should be prioritized, rather than small shifts across arbitrary borderlines. Steric criteria work best with all explicit H atoms. `Backrub' motions should be used for side chains and `P-perp' diagnostics to correct ribose puckers. A `water' may actually be an ion, a relic of misfitting or an unmodeled alternate. Beware of wishful thinking in modeling ligands. At high resolution, internally consistent alternate conformations should be modeled and geometry in poor density should not be downweighted. At low resolution, CaBLAM should be used to diagnose protein secondary structure and ERRASER to correct RNA backbone. All atoms should not be forced inside density, beware of sequence misalignment, and very rare conformations such as cis-non-Pro peptides should be avoided. Automation continues to improve, but the crystallographer still must look at each outlier, in the context of density, and correct most of them. For the valid few with unambiguous density and something that is holding them in place, a functional reason should be sought. The expectation is a few outliers, not zero.

Bioinformatics Tools and Resources for Analyzing Protein Structures.

The dramatic increase in the number of protein sequences and structures deposited in biological databases has led to the development of many bioinformatics tools and programs to manage, validate, compare, and interpret this large volume of data. In addition, powerful tools are being developed to use this sequence and structural data to facilitate protein classification and infer biological function of newly identified proteins. This chapter covers freely available bioinformatics resources on the World Wide Web that are commonly used for protein structure analysis.

Protein structure refinement via molecular-dynamics simulations: What works and what does not?

Protein structure refinement during CASP11 by the Feig group was described. Molecular dynamics simulations were used in combination with an improved selection and averaging protocol. On average, modest refinement was achieved with some targets improved significantly. Analysis of the CASP submission from our group focused on refinement success versus amount of sampling, refinement of different secondary structure elements and whether refinement varied as a function of which group provided initial models. The refinement of local stereochemical features was examined via the MolProbity score and an updated protocol was developed that can generate high-quality structures with very low MolProbity scores for most starting structures with modest computational effort. Proteins 2016; 84(Suppl 1):282-292. © 2015 Wiley Periodicals, Inc.

MolProbity for the masses-of data.

MolProbity is a powerful software program for validating structures of proteins and nucleic acids. Although MolProbity includes scripts for batch analysis of structures, because these scripts analyze structures one at a time, they are not well suited for the validation of a large dataset of structures. We have created a version of MolProbity (MolProbity-HTC) that circumvents these limitations and takes advantage of a high-throughput computing cluster by using the HTCondor software. MolProbity-HTC enables the longitudinal analysis of large sets of structures, such as those deposited in the PDB or generated through theoretical computation-tasks that would have been extremely time-consuming using previous versions of MolProbity. We have used MolProbity-HTC to validate the entire PDB, and have developed a new visual chart for the BioMagResBank website that enables users to easily ascertain the quality of each model in an NMR ensemble and to compare the quality of those models to the rest of the PDB.

Computational Methods for RNA Structure Validation and Improvement.

With increasing recognition of the roles RNA molecules and RNA/protein complexes play in an unexpected variety of biological processes, understanding of RNA structure-function relationships is of high current importance. To make clean biological interpretations from three-dimensional structures, it is imperative to have high-quality, accurate RNA crystal structures available, and the community has thoroughly embraced that goal. However, due to the many degrees of freedom inherent in RNA structure (especially for the backbone), it is a significant challenge to succeed in building accurate experimental models for RNA structures. This chapter describes the tools and techniques our research group and our collaborators have developed over the years to help RNA structural biologists both evaluate and achieve better accuracy. Expert analysis of large, high-resolution, quality-conscious RNA datasets provides the fundamental information that enables automated methods for robust and efficient error diagnosis in validating RNA structures at all resolutions. The even more crucial goal of correcting the diagnosed outliers has steadily developed toward highly effective, computationally based techniques. Automation enables solving complex issues in large RNA structures, but cannot circumvent the need for thoughtful examination of local details, and so we also provide some guidance for interpreting and acting on the results of current structure validation for RNA.