Search

An increasing number of density maps of macromolecular structures, including proteins and DNA/RNA complexes, have been determined by cryo-electron microscopy (cryo-EM). Although lately maps at a near-atomic resolution are routinely reported, there are still substantial fractions of maps determined at intermediate or low resolutions, where extracting structure information is not trivial. Here, we report a new computational method, Emap2sec+, which identifies DNA or RNA as well as the secondary structures of proteins in cryo-EM maps of 5 to 10 Å resolution. Emap2sec+ employs the deep Residual convolutional neural network. Emap2sec+ assigns structural labels with associated probabilities at each voxel in a cryo-EM map, which will help structure modeling in an EM map. Emap2sec+ showed stable and high assignment accuracy for nucleotides in low resolution maps and improved performance for protein secondary structure assignments than its earlier version when tested on simulated and experimental maps., It is challenging to extract structural information from EM density maps at intermediate or low resolutions. Here, the authors present Emap2sec+, a program for detecting nucleotides and protein secondary structures in EM density maps at 5 to 10 Å resolution.

We present the results for CAPRI Round 50, the fourth joint CASP-CAPRI protein assembly prediction challenge. The Round comprised a total of twelve targets, including six dimers, three trimers, and three higher-order oligomers. Four of these were easy targets, for which good structural templates were available either for the full assembly, or for the main interfaces (of the higher-order oligomers). Eight were difficult targets for which only distantly related templates were found for the individual subunits. Twenty-five CAPRI groups including eight automatic servers submitted ~1250 models per target. Twenty groups including six servers participated in the CAPRI scoring challenge submitted ~190 models per target. The accuracy of the predicted models was evaluated using the classical CAPRI criteria. The prediction performance was measured by a weighted scoring scheme that takes into account the number of models of acceptable quality or higher submitted by each group as part of their five top-ranking models. Compared to the previous CASP-CAPRI challenge, top performing groups submitted such models for a larger fraction (70–75%) of the targets in this Round, but fewer of these models were of high accuracy. Scorer groups achieved stronger performance with more groups submitting correct models for 70–80% of the targets or achieving high accuracy predictions. Servers performed less well in general, except for the MDOCKPP and LZERD servers, who performed on par with human groups. In addition to these results, major advances in methodology are discussed, providing an informative overview of where the prediction of protein assemblies currently stands., Cancer Research UK, Grant/Award Number: FC001003; Changzhou Science and Technology Bureau, Grant/Award Number: CE20200503; Department of Energy and Climate Change, Grant/Award Numbers: DE-AR001213, DE-SC0020400, DE-SC0021303; H2020 European Institute of Innovation and Technology, Grant/Award Numbers: 675728, 777536, 823830; Institut national de recherche en informatique et en automatique (INRIA), Grant/Award Number: Cordi-S; Lietuvos Mokslo Taryba, Grant/Award Numbers: S-MIP-17-60, S-MIP-21-35; Medical Research Council, Grant/Award Number: FC001003; Japan Society for the Promotion of Science KAKENHI, Grant/Award Number: JP19J00950; Ministerio de Ciencia e Innovación, Grant/Award Number: PID2019-110167RB-I00; Narodowe Centrum Nauki, Grant/Award Numbers: UMO-2017/25/B/ST4/01026, UMO-2017/26/M/ST4/00044, UMO-2017/27/B/ST4/00926; National Institute of General Medical Sciences, Grant/Award Numbers: R21GM127952, R35GM118078, RM1135136, T32GM132024; National Institutes of Health, Grant/Award Numbers: R01GM074255, R01GM078221, R01GM093123, R01GM109980, R01GM133840, R01GN123055, R01HL142301, R35GM124952, R35GM136409; National Natural Science Foundation of China, Grant/Award Number: 81603152; National Science Foundation, Grant/Award Numbers: AF1645512, CCF1943008, CMMI1825941, DBI1759277, DBI1759934, DBI1917263, DBI20036350, IIS1763246, MCB1925643; NWO, Grant/Award Number: TOP-PUNT 718.015.001; Wellcome Trust, Grant/Award Number: FC001003

Last year saw a breakthrough in protein structure prediction, where the AlphaFold2 method showed a substantial improvement in the modeling accuracy. Following the software release of AlphaFold2, predicted structures by AlphaFold2 for proteins in 21 species were made publicly available via the AlphaFold Database. Here, to facilitate structural analysis and application of AlphaFold2 models, we provide the infrastructure, 3D-AF-Surfer, which allows real-time structure-based search for the AlphaFold2 models. In 3D-AF-Surfer, structures are represented with 3D Zernike descriptors (3DZD), which is a rotationally invariant, mathematical representation of 3D shapes. We developed a neural network that takes 3DZDs of proteins as input and retrieves proteins of the same fold more accurately than direct comparison of 3DZDs. Using 3D-AF-Surfer, we report structure classifications of AlphaFold2 models and discuss the correlation between confidence levels of AlphaFold2 models and intrinsic disordered regions.

An increasing number of density maps of biological macromolecules have been determined by cryo-electron microscopy (cryo-EM) and stored in the public database, EMDB. To interpret the structural information contained in EM density maps, alignment of maps is an essential step for structure modeling, comparison of maps, and for database search. Here, we developed VESPER, which captures the similarity of underlying molecular structures embedded in density maps by taking local gradient directions into consideration. Compared to existing methods, VESPER achieved substantially more accurate global and local alignment of maps as well as database retrieval., Here, the authors present VESPER, a program for EM density map search and alignment. Using benchmark datasets, they demonstrate that VESPER performs accurate global and local alignments and comparisons of EM maps.

This paper describes outcomes of the 2019 Cryo-EM Model Challenge. The goals were to (1) assess the quality of models that can be produced from cryogenic electron microscopy (cryo-EM) maps using current modeling software, (2) evaluate reproducibility of modeling results from different software developers and users and (3) compare performance of current metrics used for model evaluation, particularly Fit-to-Map metrics, with focus on near-atomic resolution. Our findings demonstrate the relatively high accuracy and reproducibility of cryo-EM models derived by 13 participating teams from four benchmark maps, including three forming a resolution series (1.8 to 3.1 Å). The results permit specific recommendations to be made about validating near-atomic cryo-EM structures both in the context of individual experiments and structure data archives such as the Protein Data Bank. We recommend the adoption of multiple scoring parameters to provide full and objective annotation and assessment of the model, reflective of the observed cryo-EM map density., A multi-laboratory study in the form of a community challenge assesses the quality of models that can be produced from cryo-EM maps using different software tools, the reproducibility of models generated by different users and the performance of metrics used for model validation.

Protein structure modeling is a fundamental step for the structural interpretation of 3D electron microscopy (EM) density map. Recently, because of the significant progress of the cryo-EM technique, protein structure modeling tools are needed for EM maps determined around 4 Å resolution. At this rear atomic resolution, finding main-chain structure and assigning the amino acid sequence into EM map are still challenging problems. We have developed a de novo modeling tool named MAINMAST for EM maps at near-atomic resolution (~4.5 Å). MAINMAST can trace the backbone structure of a protein from an EM density map directory. We also developed a Graphical User Interface (GUI) plugin of MAINMAST for the UCSF Chimera so that users can monitor structures at each step of a modeling procedure. In this chapter, we demonstrate two examples of the use of MAINMAST software and MAINMAST-GUI to build protein structure model from an EM density map. MAINMAST software and MAINMAST-GUI plugin are freely available for academic users at http://kiharalab.org/mainmast/index.html .

This paper describes outcomes of the 2019 Cryo-EM Map-based Model Metrics Challenge sponsored by EMDataResource (www.emdataresource.org). The goals of this challenge were (1) to assess the quality of models that can be produced using current modeling software, (2) to check the reproducibility of modeling results from different software developers and users, and (3) compare the performance of current metrics used for evaluation of models. The focus was on near-atomic resolution maps with an innovative twist: three of four target maps formed a resolution series (1.8 to 3.1 Å) from the same specimen and imaging experiment. Tools developed in previous challenges were expanded for managing, visualizing and analyzing the 63 submitted coordinate models, and several novel metrics were introduced. The results permit specific recommendations to be made about validating near-atomic cryo-EM structures both in the context of individual laboratory experiments and holdings of structure data archives such as the Protein Data Bank. Our findings demonstrate the relatively high accuracy and reproducibility of cryo-EM models derived from these benchmark maps by 13 participating teams, representing both widely used and novel modeling approaches. We also evaluate the pros and cons of the commonly used metrics to assess model quality and recommend the adoption of multiple scoring parameters to provide full and objective annotation and assessment of the model, reflective of the observed density in the cryo-EM map.

Protein tertiary structure modeling is a critical step for the interpretation of three dimensional (3D) election microscopy density. Our group participated the 2015/2016 EM Model Challenge using the MAINMAST software for a de novo main chain modeling. The software generates local dense points using the mean shifting algorithm, and connects them into Cα models by calculating the minimum spanning tree and the longest path. Subsequently, full atom structure models are generated, which are subject to structural refinement. Here, we summarize the qualities of our submitted models and examine successful and unsuccessful models, including 3D models we did not submit to the Challenge. Our protocol using the MAINMAST software was sometimes able to build correct conformations with 3.4–5.1 A RMSD. Unsuccessful models had failure of chain traces, however, their Cα positions and some local structures were quite correctly built. For evaluate the quality of the models, the MAINMAST software provides a confidence score for each Cα position from the consensus of top 100 scoring models.

Many important functions in a cell are carried out by protein complexes with more than two subunits. Similar to the folding of a single protein, multimeric protein complexes in general follow an energetically favored assembly path. Knowing the assembly path not only provides critical information about the molecular mechanism of the assembly but also serves as a foundation for artificial design of protein complexes, as well as development of drugs that interfere with complex formation. There are experimental approaches for determining the assembly path of a complex; however, such methods are resource intensive. We have recently developed a computational method, Path-LZerD, which predicts the assembly path of a complex by simulating the docking process of the complex. Here, we explain how to use the Path-LZerD software with examples.

We report the performance of the protein docking prediction pipeline of our group and the results for Critical Assessment of Prediction of Interactions (CAPRI) rounds 38-46. The pipeline integrates programs developed in our group as well as other existing scoring functions. The core of the pipeline is the LZerD protein-protein docking algorithm. If templates of the target complex are not found in PDB, the first step of our docking prediction pipeline is to run LZerD for a query protein pair. Meanwhile, in the case of human group prediction, we survey the literature to find information that can guide the modeling, such as protein-protein interface information. In addition to any literature information and binding residue prediction, generated docking decoys were selected by a rank aggregation of statistical scoring functions. The top 10 decoys were relaxed by a short molecular dynamics simulation before submission to remove atom clashes and improve side-chain conformations. In these CAPRI rounds, our group, particularly the LZerD server, showed robust performance. On the other hand, there are failed cases where some other groups were successful. To understand weaknesses of our pipeline, we analyzed sources of errors for failed targets. Since we noted that structure refinement is a step that needs improvement, we newly performed a comparative study of several refinement approaches. Finally, we show several examples that illustrate successful and unsuccessful cases by our group.

We report the performance of protein-protein docking predictions by our group for recent rounds of the Critical Assessment of Prediction of Interactions (CAPRI), a community-wide assessment of state-of-the-art docking methods. Our prediction procedure uses a protein-protein docking program named LZerD developed in our group. LZerD represents a protein surface with 3D Zernike descriptors (3DZD), which are based on a mathematical series expansion of a 3D function. The appropriate soft representation of protein surface with 3DZD makes the method more tolerant to conformational change of proteins upon docking, which adds an advantage for unbound docking. Docking was guided by interface residue prediction performed with BindML and cons-PPISP as well as literature information when available. The generated docking models were ranked by a combination of scoring functions, including PRESCO, which evaluates the native-likeness of residues' spatial environments in structure models. First, we discuss the overall performance of our group in the CAPRI prediction rounds and investigate the reasons for unsuccessful cases. Then, we examine the performance of several knowledge-based scoring functions and their combinations for ranking docking models. It was found that the quality of a pool of docking models generated by LZerD, that is whether or not the pool includes near-native models, can be predicted by the correlation of multiple scores. Although the current analysis used docking models generated by LZerD, findings on scoring functions are expected to be universally applicable to other docking methods. Proteins 2017; 85:513-527. © 2016 Wiley Periodicals, Inc.

An increasing number of protein structures are determined by cryo-electron microscopy (cryo-EM) at near atomic resolution. However, tracing the main-chains and building full-atom models from EM maps of ~4–5 Å is still not trivial and remains a time-consuming task. Here, we introduce a fully automated de novo structure modeling method, MAINMAST, which builds three-dimensional models of a protein from a near-atomic resolution EM map. The method directly traces the protein’s main-chain and identifies Cα positions as tree-graph structures in the EM map. MAINMAST performs significantly better than existing software in building global protein structure models on data sets of 40 simulated density maps at 5 Å resolution and 30 experimentally determined maps at 2.6–4.8 Å resolution. In another benchmark of building missing fragments in protein models for EM maps, MAINMAST builds fragments of 11–161 residues long with an average RMSD of 2.68 Å., Main-chain tracing remains a time-consuming task for medium resolution cryo-EM maps. Here the authors describe MAINMAST, a computational approach for building main-chain structure models of proteins from EM maps of 4-5 Å resolution that builds main-chain models of the protein by tracing local dense points in the density distribution.

Protein-protein interactions are the cornerstone of numerous biological processes. Although an increasing number of protein complex structures have been determined using experimental methods, relatively fewer studies have been performed to determine the assembly order of complexes. In addition to the insights into the molecular mechanisms of biological function provided by the structure of a complex, knowing the assembly order is important for understanding the process of complex formation. Assembly order is also practically useful for constructing subcomplexes as a step toward solving the entire complex experimentally, designing artificial protein complexes, and developing drugs that interrupt a critical step in the complex assembly. There are several experimental methods for determining the assembly order of complexes; however, these techniques are resource-intensive. Here, we present a computational method that predicts the assembly order of protein complexes by building the complex structure. The method, named Path-LzerD, uses a multimeric protein docking algorithm that assembles a protein complex structure from individual subunit structures and predicts assembly order by observing the simulated assembly process of the complex. Benchmarked on a dataset of complexes with experimental evidence of assembly order, Path-LZerD was successful in predicting the assembly pathway for the majority of the cases. Moreover, when compared with a simple approach that infers the assembly path from the buried surface area of subunits in the native complex, Path-LZerD has the strong advantage that it can be used for cases where the complex structure is not known. The path prediction accuracy decreased when starting from unbound monomers, particularly for larger complexes of five or more subunits, for which only a part of the assembly path was correctly identified. As the first method of its kind, Path-LZerD opens a new area of computational protein structure modeling and will be an indispensable approach for studying protein complexes., Author summary Protein-protein interactions, particularly those involving multiple proteins, are the cornerstone of numerous biological processes. Although an increasing number of multi-chain protein complex structures have been determined, fewer studies have been performed to determine the assembly order of complexes. Knowing the assembly order of a complex provides insights into the process of complex formation. Assembly order is also practically useful for reconstructing and determining the structure of a subcomplex of a large protein complex. It also has important applications including designing artificial protein complexes and drugs that prevent the assembly of protein complexes. We present a computational method, Path-LZerD, which predicts the assembly order of a protein complex by simulating its assembly process. This is the first method of this kind. A strong advantage of Path-LZerD is that the assembly order can be predicted even when the overall complex structure is not known. Path-LZerD opens a new area of computational protein structure modeling and will be an indispensable approach for studying protein complexes.

Although structures determined at near-atomic resolution are now routinely reported by cryo-electron microscopy (cryo-EM), many density maps are determined at an intermediate resolution, and extracting structure information from these maps is still a challenge. We report a computational method, Emap2sec, that identifies the secondary structures of proteins (α-helices, β-sheets and other structures) in EM maps at resolutions of between 5 and 10 A. Emap2sec uses a three-dimensional deep convolutional neural network to assign secondary structure to each grid point in an EM map. We tested Emap2sec on EM maps simulated from 34 structures at resolutions of 6.0 and 10.0 A, as well as on 43 maps determined experimentally at resolutions of between 5.0 and 9.5 A. Emap2sec was able to clearly identify the secondary structures in many maps tested, and showed substantially better performance than existing methods. Emap2sec uses a deep convolutional neural network to assign protein secondary structures in intermediate-resolution (5–10 A) cryo-EM maps.

In protein structure prediction, such as template-based modeling and free modeling (ab initio modeling), the step that assesses the quality of protein models is very important. We have developed a model quality assessment (QA) program United3D that uses an optimized clustering method and a simple Cα atom contact-based potential. United3D automatically estimates the quality scores (Qscore) of predicted protein models that are highly correlated with the actual quality (GDT_TS). The performance of United3D was tested in the ninth Critical Assessment of protein Structure Prediction (CASP9) experiment. In CASP9, United3D showed the lowest average loss of GDT_TS (5.3) among the QA methods participated in CASP9. This result indicates that the performance of United3D to identify the high quality models from the models predicted by CASP9 servers on 116 targets was best among the QA methods that were tested in CASP9. United3D also produced high average Pearson correlation coefficients (0.93) and acceptable Kendall rank correlation coefficients (0.68) between the Qscore and GDT_TS. This performance was competitive with the other top ranked QA methods that were tested in CASP9. These results indicate that United3D is a useful tool for selecting high quality models from many candidate model structures provided by various modeling methods. United3D will improve the accuracy of protein structure prediction.

We have devised a power function (PF) that can predict the accuracy of a three-dimensional (3D) structure model of a protein using only amino acid sequence alignments. This Power Function (PF) consists of three parts; (1) the length of a model, (2) a homology identity percent value and (3) the agreement rate between PSI-PRED secondary structure prediction and the secondary structure judgment of a reference protein. The PF value is mathematically computed from the execution process of homology search tools, such as FASTA or various BLAST programs, to obtain the amino acid sequence alignments. There is a high correlation between the global distance test-total score (GDT_TS) value of the protein model based upon the PF score and the GDT_TS(MAX) value used as an index of protein modeling accuracy in the international contest Critical Assessment of Techniques for Protein Structure Prediction (CASP). Accordingly, the PF method is valuable for constructing a highly accurate model without wasteful calculations of homology modeling that is normally performed by an iterative method to move the main chain and side chains in the modeling process. Moreover, a model with higher accuracy can be obtained by combining the models ordered by the PF score with models sorted by the size of the CIRCLE score. The CIRCLE software is a 3D-1D program, in which energetic stabilization is estimated based upon the experimental environment of each amino acid residue in the protein solution or protein crystals.

Almost all proteins express their biological functions through the structural conformation of their specific amino acid sequences. Therefore, acquiring the three-dimensional structures of proteins is very important to elucidate the role of a particular protein. We had built protein structure model databases, which is called RIKEN FAMSBASE (http://famshelp.gsc.riken.jp/famsbase/). The RIKEN FAMSBASE is a genome-wide protein structure model database that contains a large number of protein models from many organisms. The HUMAN FAMSBASE that is one part of the RIKEN FAMSBASE contains many protein models for human genes, which are significant in the pharmaceutical and medicinal fields. We have now implemented an update of the human protein modeling database consisting of 242918 constructed models against the number of 20743 human protein sequences with an improved modeling method called Full Automatic protein Modeling System Developed (FAMSD). The results of our benchmark test of the FAMSD method indicated that it has an excellent capability to pack amino acid side-chains with correct torsion angles in addition to the main-chain, while avoiding the formation of atom-atom collisions that are not found in experimental structures. This new protein structure model database for human genes, which is named HUMAN FAMSD-BASE, is open to the public as a component part of the RIKEN FAMSBASE at http://mammalia.gsc.riken.jp/human_famsd/. A significant improvement of the HUMAN FAMSD-BASE in comparison with the preceding HUMAN FAMSBASE was verified in the benchmark test of this paper. The HUMAN FAMSD-BASE will have an important impact on the progress of biological science.

We participated in rounds 6-12 of the critical assessment of predicted interaction (CAPRI) contest as the SKE-DOCK server and human teams. The SKE-DOCK server is based on simple geometry docking and a knowledge base scoring function. The procedure is summarized in the following three steps: (1) protein docking according to shape complementarity, (2) evaluating complex models, and (3) repacking side-chain of models. The SKE-DOCK server did not make use of biological information. On the other hand, the human team tried various intervention approaches. In this article, we describe in detail the processes of the SKE-DOCK server, together with results and reasons for success and failure. Good predicted models were obtained for target 25 by both the SKE-DOCK server and human teams. When the modeled receptor proteins were superimposed on the experimental structures, the smallest Ligand-rmsd values corresponding to the rmsd between the model and experimental structures were 3.307 and 3.324 A, respectively. Moreover, the two teams obtained 4 and 2 acceptable models for target 25. The overall result for both the SKE-DOCK server and human teams was medium accuracy for one (Target 25) out of nine targets.

In CASP6, the CHIMERA-group predicted full-atom models of all targets using SKE-CHIMERA, a Web-user interface system for protein structure prediction that allows human intervention at necessary stages; we used a lot of information from our own data and from publicly available data. Using SKE-CHIMERA, we iterated manual step (template selection and alignment by the in-house program CHIMERA) and automatic step (three-dimensional model building by the in-house program FAMS). The official CASP6 assessment showed that CHIMERA-group was one of the most successful predictors in homology modeling, especially for FR/H (Fold Recognition/Homologous). In this article, we introduce the method of CHIMERA-group and discuss its successes and failures in CASP6.

Searching for protein structure-function relationships using three-dimensional (3D) structural coordinates represents a fundamental approach for determining the function of proteins with unknown functions. Since protein structure databases are rapidly growing in size, the development of a fast search method to find similar protein substructures by comparison of protein 3D structures is essential. In this article, we present a novel protein 3D structure search method to find all substructures with root mean square deviations (RMSDs) to the query structure that are lower than a given threshold value. Our new algorithm runs in O(m + N/m0.5) time, after O(N log N) preprocessing, where N is the database size and m is the query length. The new method is 1.8–41.6 times faster than the practically best known O(N) algorithm, according to computational experiments using a huge database (i.e., >20,000,000 C-alpha coordinates).

Proteins are flexible, and this flexibility has an essential functional role. Flexibility can be observed in loop regions, rearrangements between secondary structure elements, and conformational changes between entire domains. However, most protein structure alignment methods treat protein structures as rigid bodies. Thus, these methods fail to identify the equivalences of residue pairs in regions with flexibility. In this study, we considered that the evolutionary relationship between proteins corresponds directly to the residue-residue physical contacts rather than the three-dimensional (3D) coordinates of proteins. Thus, we developed a new protein structure alignment method, contact area-based alignment (CAB-align), which uses the residue-residue contact area to identify regions of similarity. The main purpose of CAB-align is to identify homologous relationships at the residue level between related protein structures. The CAB-align procedure comprises two main steps: First, a rigid-body alignment method based on local and global 3D structure superposition is employed to generate a sufficient number of initial alignments. Then, iterative dynamic programming is executed to find the optimal alignment. We evaluated the performance and advantages of CAB-align based on four main points: (1) agreement with the gold standard alignment, (2) alignment quality based on an evolutionary relationship without 3D coordinate superposition, (3) consistency of the multiple alignments, and (4) classification agreement with the gold standard classification. Comparisons of CAB-align with other state-of-the-art protein structure alignment methods (TM-align, FATCAT, and DaliLite) using our benchmark dataset showed that CAB-align performed robustly in obtaining high-quality alignments and generating consistent multiple alignments with high coverage and accuracy rates, and it performed extremely well when discriminating between homologous and nonhomologous pairs of proteins in both single and multi-domain comparisons. The CAB-align software is freely available to academic users as stand-alone software at http://www.pharm.kitasato-u.ac.jp/bmd/bmd/Publications.html.

In CAPRI Rounds 1 and 2, we assumed that because there are many ionic charges that weaken electrostatic interaction forces in living cells, the hydrophobic interaction force might be important entropically. As a result of Rounds 1 and 2, the predictions for binding sites and geometric centers were acceptable, but those of the binding axes were poor, because only the largest benzene cluster was used for generating the initial docking structures. These were generated by fitting of benzene clusters formed on the surface of receptor and ligand. In CAPRI Rounds 3-5, the grid-scoring sum on the protein-protein interaction surface and the pairwise potential of the amino acid residues, which were indicated as coming easily into the protein-protein interaction regions, were used as the calculation methods, along with the smaller benzene clusters that participated in benzene cluster fitting. Good predicted models were obtained for Targets 11 and 12. When the modeled receptor proteins were superimposed on the experimental structures, the smallest ligand root-mean-square deviation (RMSD) values corresponding to the RMSD between the model and experimental structures were 6.2 A and 7.3 A, respectively.