Verification: model-free phasing with enhanced predicted models in ARCIMBOLDO_SHREDDER

Abstract

Structure predictions have matched the accuracy of experimental structures from close homologues, providing suitable models for molecular replacement phasing. Even in predictions that present large differences due to the relative movement of domains or poorly predicted areas, very accurate regions tend to be present. These are suitable for successful fragment-based phasing as implemented in ARCIMBOLDO. The particularities of predicted models are inherently addressed in the new predicted_model mode, rendering preliminary treatment superfluous but also harmless. B-value conversion from predicted LDDT or error estimates, the removal of unstructured polypeptide, hierarchical decomposition of structural units from domains to local folds and systematically probing the model against the experimental data will ensure the optimal use of the model in phasing. Concomitantly, the exhaustive use of models and stereochemistry in phasing, refinement and validation raises the concern of crystallographic model bias and the need to critically establish the information contributed by the experiment. Therefore, in its predicted_model mode ARCIMBOLDO_SHREDDER will first determine whether the input model already constitutes a solution or provides a straightforward solution with Phaser. If not, extracted fragments will be located. If the landscape of solutions reveals numerous, clearly discriminated and consistent probes or if the input model already constitutes a solution, model-free verification will be activated. Expansions with SHELXE will omit the partial solution seeding phases and all traces outside their respective masks will be combined in ALIXE, as far as consistent. This procedure completely eliminates the molecular replacement search model in favour of the inferences derived from this model. In the case of fragments, an incorrect starting hypothesis impedes expansion. The predicted_model mode has been tested in different scenarios.

The CCP4 suite: integrative software for macromolecular crystallography

“…CCP4 also has fragment-based ab initio phasing packages: ARCIMBOLDO (Rodrı ´guez et al, 2009) and Fragon (Jenkins, 2018), which use ideal fragments of proteins (mainly helices) in targeted molecular-replacement searches. The use of these programs was initially confined to high-resolution data, but they have recently enjoyed success at resolutions lower than 2.3 A ˚, a threshold beyond which it becomes difficult to ascertain the direction of helical fragments, owing to their improved search strategies (Medina et al, 2022), phase combination (Milla ´n et al, 2020) and the use of available structural information, including AlphaFold predictions. ARCIMBOLDO (Rodrı ´guez et al, 2009) can use fragments of homologous models and phase previously intractable coiledcoil structures (Caballero et al, 2018).…”

“…The last decade has seen some large transformations in the field of MX: new workflows have been created (for example phasing with AlphaFold2 models) and some old workflows have been optimized, while some others are on the verge of disappearing; this has often been the result of cross-pollination with other techniques in structural biology, for example electron cryo-microscopy (cryo-EM) in particular, through a synergistic collaboration with CCP-EM (Burnley et al, 2017), the Collaborative Computational Project for Cryo-EM, which repurposes some CCP4 code for the cryo-EM community. For example, owing to the deep-learning revolution in computational structure prediction (Jumper et al, 2021), it is now possible to phase most structures using large predicted fragments or, owing to the accuracy of the method, even to rigidbody fit an initial predicted model into electron density (Oeffner et al, 2022;McCoy et al, 2022;Medina et al, 2022). As a side effect of the creation of these new workflows, experimental phasing is now losing importance in the everyday activities of an MX laboratory, with derivatives only being created as a last resort after all of the now conventional methods have failed.…”

Predicted models and CCP4

“…In the case where the predicted structure is a coiled coil, the user should select the coiled_coil mode along with the predicted_model mode through the interfaces; this will replace the model-free verification with a dedicated verification addressing the specific pitfalls arising from the modulation and anisotropy (Caballero et al, 2018). Finally, if the structure is a multimer and expansion of a first placement does not suffice to provide a solution, the multicopy approach will be activated to sequentially search for several copies with an optimized prioritization step to speed up calculations (Medina et al, 2022). This procedure is illustrated in Fig.…”

“…The ARCIMBOLDO_ SHREDDER sequential and ARCIMBOLDO_SHREDDER spheres programs (Sammito et al, 2014;Milla ´n et al, 2018) were originally developed for phasing using fragments that were extracted from remote homologs, identified and refined against the experimental data. ARCIMBOLDO_SHREDDER spheres was thus well suited to solve structures with any kind of structural template and the method has been adapted to optimize the use of predicted models, while systematically removing model bias (Medina et al, 2022).…”

Accelerating crystal structure determination with iterative AlphaFold prediction

“…The potential for using AlphaFold predictions to facilitate structure determination by X-ray crystallography and cryo-EM has been rapidly appreciated in the structural biology community (Akdel et al, 2022;Barbarin-Bocahu & Graille, 2022;Bond & Cowtan, 2022;Chen et al, 2022;Gong et al, 2023;McCoy et al, 2022;Medina et al, 2022;Moi et al, 2022;Stsiapanava et al, 2022). AlphaFold predictions made in the CASP14 blind test of structure prediction were shown to be effective as starting models for X-ray structure determination using the molecular-replacement method (McCoy et al, 2022).…”