AlphaFold predictions are valuable hypotheses and accelerate but do not replace experimental structure determination.

Artificial intelligence-based protein structure prediction methods such as AlphaFold have revolutionized structural biology. The accuracies of these predictions vary, however, and they do not take into account ligands, covalent modifications or other environmental factors. Here, we evaluate how well AlphaFold predictions can be expected to describe the structure of a protein by comparing predictions directly with experimental crystallographic maps. In many cases, AlphaFold predictions matched experimental maps remarkably closely. In other cases, even very high-confidence predictions differed from experimental maps on a global scale through distortion and domain orientation, and on a local scale in backbone and side-chain conformation. We suggest considering AlphaFold predictions as exceptionally useful hypotheses. We further suggest that it is important to consider the confidence in prediction when interpreting AlphaFold predictions and to carry out experimental structure determination to verify structural details, particularly those that involve interactions not included in the prediction.

Improved AlphaFold modeling with implicit experimental information.

Machine-learning prediction algorithms such as AlphaFold and RoseTTAFold can create remarkably accurate protein models, but these models usually have some regions that are predicted with low confidence or poor accuracy. We hypothesized that by implicitly including new experimental information such as a density map, a greater portion of a model could be predicted accurately, and that this might synergistically improve parts of the model that were not fully addressed by either machine learning or experiment alone. An iterative procedure was developed in which AlphaFold models are automatically rebuilt on the basis of experimental density maps and the rebuilt models are used as templates in new AlphaFold predictions. We show that including experimental information improves prediction beyond the improvement obtained with simple rebuilding guided by the experimental data. This procedure for AlphaFold modeling with density has been incorporated into an automated procedure for interpretation of crystallographic and electron cryo-microscopy maps.

Sizes of pure and doped helium droplets from single shot x-ray imaging.

Advancements in x-ray free-electron lasers on producing ultrashort, ultrabright, and coherent x-ray pulses enable single-shot imaging of fragile nanostructures, such as superfluid helium droplets. This imaging technique gives unique access to the sizes and shapes of individual droplets. In the past, such droplet characteristics have only been indirectly inferred by ensemble averaging techniques. Here, we report on the size distributions of both pure and doped droplets collected from single-shot x-ray imaging and produced from the free-jet expansion of helium through a 5 µm diameter nozzle at 20 bars and nozzle temperatures ranging from 4.2 to 9 K. This work extends the measurement of large helium nanodroplets containing 109-1011 atoms, which are shown to follow an exponential size distribution. Additionally, we demonstrate that the size distributions of the doped droplets follow those of the pure droplets at the same stagnation condition but with smaller average sizes.

The Phenix-AlphaFold webservice: Enabling AlphaFold predictions for use in Phenix.

Advances in machine learning have enabled sufficiently accurate predictions of protein structure to be used in macromolecular structure determination with crystallography and cryo-electron microscopy data. The Phenix software suite has AlphaFold predictions integrated into an automated pipeline that can start with an amino acid sequence and data, and automatically perform model-building and refinement to return a protein model fitted into the data. Due to the steep technical requirements of running AlphaFold efficiently, we have implemented a Phenix-AlphaFold webservice that enables all Phenix users to run AlphaFold predictions remotely from the Phenix GUI starting with the official 1.21 release. This webservice will be improved based on how it is used by the research community and the future research directions for Phenix.

In situ ligand restraints from quantum-mechanical methods.

In macromolecular crystallographic structure refinement, ligands present challenges for the generation of geometric restraints due to their large chemical variability, their possible novel nature and their specific interaction with the binding pocket of the protein. Quantum-mechanical approaches are useful for providing accurate ligand geometries, but can be plagued by the number of minima in flexible molecules. In an effort to avoid these issues, the Quantum Mechanical Restraints (QMR) procedure optimizes the ligand geometry in situ, thus accounting for the influence of the macromolecule on the local energy minima of the ligand. The optimized ligand geometry is used to generate target values for geometric restraints during the crystallographic refinement. As demonstrated using a sample of >2330 ligand instances in >1700 protein-ligand models, QMR restraints generally result in lower deviations from the target stereochemistry compared with conventionally generated restraints. In particular, the QMR approach provides accurate torsion restraints for ligands and other entities.

Conformational space exploration of cryo-EM structures by variability refinement.

Cryo-EM observation of biological samples enables visualization of sample heterogeneity, in the form of discrete states that are separable, or continuous heterogeneity as a result of local protein motion before flash freezing. Variability analysis of this continuous heterogeneity describes the variance between a particle stack and a volume, and results in a map series describing the various steps undertaken by the sample in the particle stack. While this observation is absolutely stunning, it is very hard to pinpoint structural details to elements of the maps. In order to bridge the gap between observation and explanation, we designed a tool that refines an ensemble of structures into all the maps from variability analysis. Using this bundle of structures, it is easy to spot variable parts of the structure, as well as the parts that are not moving. Comparison with molecular dynamics simulations highlights the fact that the movements follow the same directions, albeit with different amplitudes. Ligand can also be investigated using this method. Variability refinement is available in the Phenix software suite, accessible under the program name phenix.varref.

Improved joint X-ray and neutron refinement procedure in Phenix.

Neutron diffraction is one of the three crystallographic techniques (X-ray, neutron and electron diffraction) used to determine the atomic structures of molecules. Its particular strengths derive from the fact that H (and D) atoms are strong neutron scatterers, meaning that their positions, and thus protonation states, can be derived from crystallographic maps. However, because of technical limitations and experimental obstacles, the quality of neutron diffraction data is typically much poorer (completeness, resolution and signal to noise) than that of X-ray diffraction data for the same sample. Further, refinement is more complex as it usually requires additional parameters to describe the H (and D) atoms. The increase in the number of parameters may be mitigated by using the `riding hydrogen' refinement strategy, in which the positions of H atoms without a rotational degree of freedom are inferred from their neighboring heavy atoms. However, this does not address the issues related to poor data quality. Therefore, neutron structure determination often relies on the presence of an X-ray data set for joint X-ray and neutron (XN) refinement. In this approach, the X-ray data serve to compensate for the deficiencies of the neutron diffraction data by refining one model simultaneously against the X-ray and neutron data sets. To be applicable, it is assumed that both data sets are highly isomorphous, and preferably collected from the same crystals and at the same temperature. However, the approach has a number of limitations that are discussed in this work by comparing four separately re-refined neutron models. To address the limitations, a new method for joint XN refinement is introduced that optimizes two different models against the different data sets. This approach is tested using neutron models and data deposited in the Protein Data Bank. The efficacy of refining models with H atoms as riding or as individual atoms is also investigated.

Interpreting macromolecular diffraction through simulation.

This chapter discusses the use of diffraction simulators to improve experimental outcomes in macromolecular crystallography, in particular for future experiments aimed at diffuse scattering. Consequential decisions for upcoming data collection include the selection of either a synchrotron or free electron laser X-ray source, rotation geometry or serial crystallography, and fiber-coupled area detector technology vs. pixel-array detectors. The hope is that simulators will provide insights to make these choices with greater confidence. Simulation software, especially those packages focused on physics-based calculation of the diffraction, can help to predict the location, size, shape, and profile of Bragg spots and diffuse patterns in terms of an underlying physical model, including assumptions about the crystal's mosaic structure, and therefore can point to potential issues with data analysis in the early planning stages. Also, once the data are collected, simulation may offer a pathway to improve the measurement of diffraction, especially with weak data, and might help to treat problematic cases such as overlapping patterns.

Accelerating crystal structure determination with iterative AlphaFold prediction.

Experimental structure determination can be accelerated with artificial intelligence (AI)-based structure-prediction methods such as AlphaFold. Here, an automatic procedure requiring only sequence information and crystallographic data is presented that uses AlphaFold predictions to produce an electron-density map and a structural model. Iterating through cycles of structure prediction is a key element of this procedure: a predicted model rebuilt in one cycle is used as a template for prediction in the next cycle. This procedure was applied to X-ray data for 215 structures released by the Protein Data Bank in a recent six-month period. In 87% of cases our procedure yielded a model with at least 50% of Cα atoms matching those in the deposited models within 2 Å. Predictions from the iterative template-guided prediction procedure were more accurate than those obtained without templates. It is concluded that AlphaFold predictions obtained based on sequence information alone are usually accurate enough to solve the crystallographic phase problem with molecular replacement, and a general strategy for macromolecular structure determination that includes AI-based prediction both as a starting point and as a method of model optimization is suggested.

Structure determination of Pt-coated Au dumbbells via fluctuation X-ray scattering.

A fluctuation X-ray scattering experiment has been carried out on platinum-coated gold nanoparticles randomly oriented on a substrate. A complete algorithm for determining the electron density of an individual particle from diffraction patterns of many particles randomly oriented about a single axis is demonstrated. This algorithm operates on angular correlations among the measured intensity distributions and recovers the angular correlation functions of a single particle from measured diffraction patterns. Taking advantage of the cylindrical symmetry of the nanoparticles, a cylindrical slice model is proposed to reconstruct the structure of the nanoparticles by fitting the experimental ring angular auto-correlation and small-angle scattering data obtained from many scattering patterns. The physical meaning of the refined structure is discussed in terms of their statistical distributions of the shape and electron density profile.

Putting AlphaFold models to work with phenix.process_predicted_model and ISOLDE.

AlphaFold has recently become an important tool in providing models for experimental structure determination by X-ray crystallography and cryo-EM. Large parts of the predicted models typically approach the accuracy of experimentally determined structures, although there are frequently local errors and errors in the relative orientations of domains. Importantly, residues in the model of a protein predicted by AlphaFold are tagged with a predicted local distance difference test score, informing users about which regions of the structure are predicted with less confidence. AlphaFold also produces a predicted aligned error matrix indicating its confidence in the relative positions of each pair of residues in the predicted model. The phenix.process_predicted_model tool downweights or removes low-confidence residues and can break a model into confidently predicted domains in preparation for molecular replacement or cryo-EM docking. These confidence metrics are further used in ISOLDE to weight torsion and atom-atom distance restraints, allowing the complete AlphaFold model to be interactively rearranged to match the docked fragments and reducing the need for the rebuilding of connecting regions.

CERES: a cryo-EM re-refinement system for continuous improvement of deposited models.

The field of electron cryomicroscopy (cryo-EM) has advanced quickly in recent years as the result of numerous technological and methodological developments. This has led to an increase in the number of atomic structures determined using this method. Recently, several tools for the analysis of cryo-EM data and models have been developed within the Phenix software package, such as phenix.real_space_refine for the refinement of atomic models against real-space maps. Also, new validation metrics have been developed for low-resolution cryo-EM models. To understand the quality of deposited cryo-EM structures and how they might be improved, models deposited in the Protein Data Bank that have map resolutions of better than 5 Šwere automatically re-refined using current versions of Phenix tools. The results are available on a publicly accessible web page (https://cci.lbl.gov/ceres). The implementation of a Cryo-EM Re-refinement System (CERES) for the improvement of models deposited in the wwPDB, and the results of the re-refinements, are described. Based on these results, contents are proposed for a `cryo-EM Table 1', which summarizes experimental details and validation metrics in a similar way to `Table 1' in crystallography. The consistent use of robust metrics for the evaluation of cryo-EM models and data should accompany every structure deposition and be reported in scientific publications.

Detection and correction of underassigned rotational symmetry prior to structure deposition.

Up to 2% of X-ray structures in the Protein Data Bank (PDB) potentially fit into a higher symmetry space group. Redundant protein chains in these structures can be made compatible with exact crystallographic symmetry with minimal atomic movements that are smaller than the expected range of coordinate uncertainty. The incidence of problem cases is somewhat difficult to define precisely, as there is no clear line between underassigned symmetry, in which the subunit differences are unsupported by the data, and pseudosymmetry, in which the subunit differences rest on small but significant intensity differences in the diffraction pattern. To help catch symmetry-assignment problems in the future, it is useful to add a validation step that operates on the refined coordinates just prior to structure deposition. If redundant symmetry-related chains can be removed at this stage, the resulting model (in a higher symmetry space group) can readily serve as an isomorphous replacement starting point for re-refinement using re-indexed and re-integrated raw data. These ideas are implemented in new software tools available at http://cci.lbl.gov/labelit.

Application of normal-mode refinement to X-ray crystal structures at the lower resolution limit.

The structural refinement of large complexes at the lower resolution limit is often difficult and inefficient owing to the limited number of reflections and the frequently high-level structural flexibility. A new normal-mode-based X-ray crystallographic refinement method has recently been developed that enables anisotropic B-factor refinement using a drastically smaller number of thermal parameters than even isotropic refinement. Here, the method has been systematically tested on a total of eight systems in the resolution range 3.0-3.9 A. This series of tests established the most applicable scenarios for the method, the detailed procedures for its application and the degree of structural improvement. The results demonstrated substantial model improvement at the lower resolution limit, especially in cases in which other methods such as the translation-libration-screw (TLS) model were not applicable owing to the poorly converged isotropic B-factor distribution. It is expected that this normal-mode-based method will be a useful tool for structural refinement, in particular at the lower resolution limit, in the field of X-ray crystallography.

Structural improvement of unliganded simian immunodeficiency virus gp120 core by normal-mode-based X-ray crystallographic refinement.

The envelope protein gp120/gp41 of simian and human immunodeficiency viruses plays a critical role in viral entry into host cells. However, the extraordinarily high structural flexibility and heavy glycosylation of the protein have presented enormous difficulties in the pursuit of high-resolution structural investigation of some of its conformational states. An unliganded and fully glycosylated gp120 core structure was recently determined to 4.0 A resolution. The rather low data-to-parameter ratio limited refinement efforts in the original structure determination. In this work, refinement of this gp120 core structure was carried out using a normal-mode-based refinement method that has been shown in previous studies to be effective in improving models of a supramolecular complex at 3.42 A resolution and of a membrane protein at 3.2 A resolution. By using only the first four nonzero lowest-frequency normal modes to construct the anisotropic thermal parameters, combined with manual adjustments and standard positional refinement using REFMAC5, the structural model of the gp120 core was significantly improved in many aspects, including substantial decreases in R factors, better fitting of several flexible regions in electron-density maps, the addition of five new sugar rings at four glycan chains and an excellent correlation of the B-factor distribution with known structural flexibility. These results further underscore the effectiveness of this normal-mode-based method in improving models of protein and nonprotein components in low-resolution X-ray structures.

Normal-mode refinement of anisotropic thermal parameters for potassium channel KcsA at 3.2 A crystallographic resolution.

We report a normal-mode method for anisotropic refinement of membrane-protein structures, based on a hypothesis that the global near-native-state disordering of membrane proteins in crystals follows low-frequency normal modes. Thus, a small set of modes is sufficient to represent the anisotropic thermal motions in X-ray crystallographic refinement. By applying the method to potassium channel KcsA at 3.2 A, we obtained a structural model with an improved fit with the diffraction data. Moreover, the improved electron density maps allowed for large structural adjustments for 12 residues in each subunit, including the rebuilding of 3 missing side chains. Overall, the anisotropic KcsA structure at 3.2 A was systematically closer to a 2.0 A KcsA structure, especially in the selectivity filter. Furthermore, the anisotropic thermal ellipsoids from the refinement revealed functionally relevant structural flexibility. We expect this method to be a valuable tool for structural refinement of many membrane proteins with moderate-resolution diffraction data.

Macromolecular structure determination using X-rays, neutrons and electrons: recent developments in Phenix.

Diffraction (X-ray, neutron and electron) and electron cryo-microscopy are powerful methods to determine three-dimensional macromolecular structures, which are required to understand biological processes and to develop new therapeutics against diseases. The overall structure-solution workflow is similar for these techniques, but nuances exist because the properties of the reduced experimental data are different. Software tools for structure determination should therefore be tailored for each method. Phenix is a comprehensive software package for macromolecular structure determination that handles data from any of these techniques. Tasks performed with Phenix include data-quality assessment, map improvement, model building, the validation/rebuilding/refinement cycle and deposition. Each tool caters to the type of experimental data. The design of Phenix emphasizes the automation of procedures, where possible, to minimize repetitive and time-consuming manual tasks, while default parameters are chosen to encourage best practice. A graphical user interface provides access to many command-line features of Phenix and streamlines the transition between programs, project tracking and re-running of previous tasks.

Interactive comparison and remediation of collections of macromolecular structures.

Often similar structures need to be compared to reveal local differences throughout the entire model or between related copies within the model. Therefore, a program to compare multiple structures and enable correction any differences not supported by the density map was written within the Phenix framework (Adams et al., Acta Cryst 2010; D66:213-221). This program, called Structure Comparison, can also be used for structures with multiple copies of the same protein chain in the asymmetric unit, that is, as a result of non-crystallographic symmetry (NCS). Structure Comparison was designed to interface with Coot(Emsley et al., Acta Cryst 2010; D66:486-501) and PyMOL(DeLano, PyMOL 0.99; 2002) to facilitate comparison of large numbers of related structures. Structure Comparison analyzes collections of protein structures using several metrics, such as the rotamer conformation of equivalent residues, displays the results in tabular form and allows superimposed protein chains and density maps to be quickly inspected and edited (via the tools in Coot) for consistency, completeness and correctness.

New tools for the analysis and validation of cryo-EM maps and atomic models.

Recent advances in the field of electron cryomicroscopy (cryo-EM) have resulted in a rapidly increasing number of atomic models of biomacromolecules that have been solved using this technique and deposited in the Protein Data Bank and the Electron Microscopy Data Bank. Similar to macromolecular crystallography, validation tools for these models and maps are required. While some of these validation tools may be borrowed from crystallography, new methods specifically designed for cryo-EM validation are required. Here, new computational methods and tools implemented in PHENIX are discussed, including d99 to estimate resolution, phenix.auto_sharpen to improve maps and phenix.mtriage to analyze cryo-EM maps. It is suggested that cryo-EM half-maps and masks should be deposited to facilitate the evaluation and validation of cryo-EM-derived atomic models and maps. The application of these tools to deposited cryo-EM atomic models and maps is also presented.

Real-space refinement in PHENIX for cryo-EM and crystallography.

This article describes the implementation of real-space refinement in the phenix.real_space_refine program from the PHENIX suite. The use of a simplified refinement target function enables very fast calculation, which in turn makes it possible to identify optimal data-restraint weights as part of routine refinements with little runtime cost. Refinement of atomic models against low-resolution data benefits from the inclusion of as much additional information as is available. In addition to standard restraints on covalent geometry, phenix.real_space_refine makes use of extra information such as secondary-structure and rotamer-specific restraints, as well as restraints or constraints on internal molecular symmetry. The re-refinement of 385 cryo-EM-derived models available in the Protein Data Bank at resolutions of 6 Šor better shows significant improvement of the models and of the fit of these models to the target maps.