Accuracy Of Structure Prediction Research Articles

A traffic prediction method for missing data scenarios: graph convolutional recurrent ordinary differential equation network

Traffic prediction plays an increasingly important role in intelligent transportation systems and smart cities. Both travelers and urban managers rely on accurate traffic information to make decisions about route selection and traffic management. Due to various factors, both human and natural, traffic data often contains missing values. Addressing the impact of missing data on traffic flow prediction has become a widely discussed topic in the academic community and holds significant practical importance. Existing spatiotemporal graph models typically rely on complete data, and the presence of missing values can significantly degrade prediction performance and disrupt the construction of dynamic graph structures. To address this challenge, this paper proposes a neural network architecture designed specifically for missing data scenarios—graph convolutional recurrent ordinary differential equation network (GCRNODE). GCRNODE combines recurrent networks based on ordinary differential equation (ODE) with spatiotemporal memory graph convolutional networks, enabling accurate traffic prediction and effective modeling of dynamic graph structures even in the presence of missing data. GCRNODE uses ODE to model the evolution of traffic flow and updates the hidden states of the ODE through observed data. Additionally, GCRNODE employs a data-independent spatiotemporal memory graph convolutional network to capture the dynamic spatial dependencies in missing data scenarios. The experimental results on three real-world traffic datasets demonstrate that GCRNODE outperforms baseline models in prediction performance under various missing data rates and scenarios. This indicates that the proposed method has stronger adaptability and robustness in handling missing data and modeling spatiotemporal dependencies.

Rational Modification of a Cross-Linker for Improved Flexible Protein Structure Modeling.

Chemical cross-linking/mass spectrometry (XL-MS) has emerged as a complementary tool for mapping interaction sites within protein networks as well as gaining moderate-resolution native structural insight with minimal interference. XL-MS technology mostly relies on chemoselective reactions (cross-linking) between protein residues and a linker. DSSO represents a versatile cross-linker for protein structure investigation and in-cell XL-MS. However, our assessment of its shelf life and batch purity revealed decomposition of DSSO in anhydrous solution via a retro-Michael reaction, which may reduce the active ingredient down to below 90%. To mitigate the occurrence of this degradative mechanism, we report the rational design and synthesis of DSSO-carbamate, which contains an inserted nitrogen atom in the DSSO backbone structure. This modification to DSSO yielded remarkably favorable stability against such decomposition, which translated to higher cross-link and monolink recovery when performing XL-MS on monomeric flexible proteins. Recently, XL-MS has been leveraged against AlphaFold2 and other protein structure prediction algorithms for improved prediction of flexible monomeric multiconformational proteins. To this end, we demonstrate that our novel cross-linker, termed DSSO-carbamate, generated more accurate protein structure predictions when combined with AlphaFold2, on account of its increased recovery of cross-links and monolinks, compared to DSSO. As such, DSSO-carbamate represents a useful addition to the XL-MS community, particularly for protein structure prediction.

Porter 6: Protein Secondary Structure Prediction by Leveraging Pre-Trained Language Models (PLMs)

Accurately predicting protein secondary structure (PSSP) is crucial for understanding protein function, which is foundational to advancements in drug development, disease treatment, and biotechnology. Researchers gain critical insights into protein folding and function within cells by predicting protein secondary structures. The advent of deep learning models, capable of processing complex sequence data and identifying meaningful patterns, offer substantial potential to enhance the accuracy and efficiency of protein structure predictions. In particular, recent breakthroughs in deep learning—driven by the integration of natural language processing (NLP) algorithms—have significantly advanced the field of protein research. Inspired by the remarkable success of NLP techniques, this study harnesses the power of pre-trained language models (PLMs) to advance PSSP prediction. We conduct a comprehensive evaluation of various deep learning models trained on distinct sequence embeddings, including one-hot encoding and PLM-based approaches such as ProtTrans and ESM-2, to develop a cutting-edge prediction system optimized for accuracy and computational efficiency. Our proposed model, Porter 6, is an ensemble of CBRNN-based predictors, leveraging the protein language model ESM-2 as input features. Porter 6 achieves outstanding performance on large-scale, independent test sets. On a 2022 test set, the model attains an impressive 86.60% accuracy in three-state (Q3) and 76.43% in eight-state (Q8) classifications. When tested on a more recent 2024 test set, Porter 6 maintains robust performance, achieving 84.56% in Q3 and 74.18% in Q8 classifications. This represents a significant 3% improvement over its predecessor, outperforming or matching state-of-the-art approaches in the field.

KnotFold: Improving Peptide Structure Predictions with Simulated Coevolution

We present a novel approach for embedding experimental contact information in the protein structure prediction tool AlphaFold2 that harnesses the principles of protein sequence coevolution to enforce disulfide connectivity patterns in the prediction of disulfide-rich peptide (DRP) structures. While AlphaFold2 generates accurate DRP structure prediction in most cases, it sometimes fails at predicting the specific connectivity pattern of the multiple disulfide bonds. Here, we take advantage of the principles of sequence coevolution to engineer the multiple sequence alignment (MSA) input with specific connectivity patterns by mutating highly conserved cysteines in subsets of the MSA—which we term “simulated coevolution.” Engineering the MSA prompt with simulated coevolution provides sufficient context in most cases for AlphaFold2 to obey the provided connectivity. Our approach—KnotFold—can be used to incorporate experimental disulfide connectivity patterns from mass spectrometry into DRP structure prediction. Lastly, we find that peptide properties and AF2Rank Composite Confidence scores can be used to differentiate native from non-native connectivity patterns in some cases, improving the ability to determine native connectivity of DRPs from sequence alone. Published by the American Physical Society 2024

Comprehensive Analysis and Application Research of Advanced Computational Algorithms in Protein Folding Simulations

Protein engineering stands at the forefront of biotechnology, aiming to modify natural proteins or create new ones tailored to specific functional requirements. The three-dimensional structures of proteins, particularly their folding patterns, are critical in defining their biological roles. Accurate prediction and detailed examination of these protein folding structures are crucial in protein engineering. The close relationship between protein structure and function highlights the importance of understanding protein folding dynamics to successfully manipulate protein designs for intended uses. Genetic algorithms (GA), taking inspiration from natural evolutionary principles, employ a heuristic search approach that integrates elements of randomness. In contrast, simulated annealing (SA) leverages stochastic optimization techniques based on the Monte Carlo method, theoretically capable of approximating the global optimum with a high degree of accuracy. Additionally, generalized ensemble methods are increasingly used to explore protein folding processes. This paper explores the fundamental principles and practical applications of these algorithms in simulating protein folding dynamics, aiming to enhance the methodologies used in protein engineering. This exploration not only aids in the refinement of protein design but also extends the potential applications of engineered proteins in various scientific and industrial fields.

Addendum: Accurate structure prediction of biomolecular interactions with AlphaFold 3

Addendum: Accurate structure prediction of biomolecular interactions with AlphaFold 3

A novel deep neural network-based technique for network embedding.

In this paper, the graph segmentation (GSeg) method has been proposed. This solution is a novel graph neural network framework for network embedding that leverages the inherent characteristics of nodes and the underlying local network topology. The key innovation of GSeg lies in its encoder-decoder architecture, which is specifically designed to preserve the network's structural properties. The key contributions of GSeg are: (1) a novel graph neural network architecture that effectively captures local and global network structures, and (2) a robust node representation learning approach that achieves superior performance in various network analysis tasks. The methodology employed in our study involves the utilization of a graph neural network framework for the acquisition of node representations. The design leverages the inherent characteristics of nodes and the underlying local network topology. To enhance the architectural framework of encoder- decoder networks, the GSeg model is specifically devised to exhibit a structural resemblance to the SegNet model. The obtained empirical results on multiple benchmark datasets demonstrate that the GSeg outperforms existing state-of-the-art methods in terms of network structure preservation and prediction accuracy for downstream tasks. The proposed technique has potential utility across a range of practical applications in the real world.

Dissecting AlphaFold2’s Capabilities with Limited Sequence Information

Abstract Protein structure prediction aims to infer a protein’s three dimensional structure from its amino acid sequence. Protein structure is pivotal for elucidating protein functions, interactions, and driving biotechnological innovation. The deep learning model AlphaFold2, has revolutionized this field by leveraging phylogenetic information from multiple sequence alignments to achieve remarkable accuracy in protein structure prediction. However, a key question remains: how well does AlphaFold2 understand protein structures? This study investigates AlphaFold2’s capabilities when relying primarily on high-quality template structures, without the additional information provided by multiple sequence alignments. By designing experiments that probe local and global structural understanding, we aimed to dissect its dependence on specific features and its ability to handle missing information. Our findings revealed AlphaFold2’s reliance on sterically valid Cβ for correctly interpreting structural templates. Additionally, we observed its remarkable ability to recover three dimensional structures from certain perturbations and the negligible impact of the previous structure in recycling. Collectively, these results support the hypothesis that AlphaFold2 has learned an accurate biophysical energy function. However, this function seems most effective for local interactions. Our work advances understanding of how deep learning models predict protein structures and provides guidance for researchers aiming to overcome limitations in these models.

Prediction of Protein Secondary Structures Based on Substructural Descriptors of Molecular Fragments.

The accurate prediction of secondary structures of proteins (SSPs) is a critical challenge in molecular biology and structural bioinformatics. Despite recent advancements, this task remains complex and demands further exploration. This study presents a novel approach to SSP prediction using atom-centric substructural multilevel neighborhoods of atoms (MNA) descriptors for protein molecular fragments. A dataset comprising over 335,000 SSPs, annotated by the Dictionary of Secondary Structure in Proteins (DSSP) software from 37,000 proteins, was constructed from Protein Data Bank (PDB) records with a resolution of 2 Å or better. Protein fragments were converted into structural formulae using the RDKit Python package and stored in SD files using the MOL V3000 format. Classification sequence-structure-property relationships (SSPR) models were developed with varying levels of MNA descriptors and a Bayesian algorithm implemented in MultiPASS software. The average prediction accuracy (AUC) for eight SSP types, calculated via leave-one-out cross-validation, was 0.902. For independent test sets (ASTRAL and CB513 datasets), the best SSPR models achieved AUC, Q3, and Q8 values of 0.860, 77.32%, 70.92% and 0.889, 78.78%, 74.74%, respectively. Based on the created models, a freely available web application MNA-PSS-Pred was developed.

Accurate RNA 3D structure prediction using a language model-based deep learning approach

Accurate prediction of RNA three-dimensional (3D) structures remains an unsolved challenge. Determining RNA 3D structures is crucial for understanding their functions and informing RNA-targeting drug development and synthetic biology design. The structural flexibility of RNA, which leads to the scarcity of experimentally determined data, complicates computational prediction efforts. Here we present RhoFold+, an RNA language model-based deep learning method that accurately predicts 3D structures of single-chain RNAs from sequences. By integrating an RNA language model pretrained on ~23.7 million RNA sequences and leveraging techniques to address data scarcity, RhoFold+ offers a fully automated end-to-end pipeline for RNA 3D structure prediction. Retrospective evaluations on RNA-Puzzles and CASP15 natural RNA targets demonstrate the superiority of RhoFold+ over existing methods, including human expert groups. Its efficacy and generalizability are further validated through cross-family and cross-type assessments, as well as time-censored benchmarks. Additionally, RhoFold+ predicts RNA secondary structures and interhelical angles, providing empirically verifiable features that broaden its applicability to RNA structure and function studies.

Heterogeneous folding landscapes and predetermined breaking points within a protein family.

The accurate prediction of protein structures with artificial intelligence has been a spectacular success. Yet, how proteins fold into their native structures inside the cell remains incompletely understood. Of particular interest is to rationalize how proteins interact with the protein homeostasis network, an organism specific set of protein folding and quality control enzymes. Failure of protein homeostasis leads to widespread misfolding and aggregation, and thus neurodegeneration. Here, I present a comparative analysis of the folding of 16 single-domain proteins from the same organism across a protein family, the Saccharomyces cerevisiae small GTPases. Using computational modeling to directly probe protein folding dynamics, this work shows how near identical structures from the same folding environment can exhibit heterogeneous folding landscapes. Remarkably, yeast small GTPases are found to unfold along different pathways either via the N- or C-terminus initiated by structure-encoded predetermined breaking points. Degrons as recognition signals for ubiquitin-dependent degradation were systematically absent from the initial unfolding sites, as if to protect from too rapid degradation upon spontaneous unfolding or before completion of the folding. The presented results highlight a direct coordination of folding pathway and protein homeostasis interaction signals across a protein family. A deeper understanding of the interdependence of proteins with their folding environment will help to rationalize and combat diseases linked to protein misfolding and dysregulation. More generally, this work underlines the importance of understanding protein folding in the cellular context, and highlights valuable constraints towards a systems-level understanding of protein homeostasis.

Protein Design by Integrating Machine Learning and Quantum-Encoded Optimization

The protein design problem involves finding polypeptide sequences folding into a given three-dimensional structure. Its rigorous algorithmic solution is computationally demanding, involving a nested search in sequence and structure spaces. Structure searches can now be bypassed thanks to recent machine-learning breakthroughs, which have enabled accurate and rapid structure predictions. Similarly, sequence searches might be entirely transformed by the advent of quantum annealing machines and by the required new encodings of the search problem, which could be performative even on classical machines. In this work, we introduce a general protein design scheme where algorithmic and technological advancements in machine learning and quantum-inspired algorithms can be integrated, and an optimal physics-based scoring function is iteratively learned. In this first proof-of-concept application, we apply the iterative method to a lattice protein model amenable to exhaustive benchmarks, finding that it can rapidly learn a physics-based scoring function and achieve promising design performances. Strikingly, our quantum-inspired reformulation outperforms conventional sequence optimization even when adopted on classical machines. The scheme is general and can be extended, e.g., to encompass off-lattice models, and it can integrate progress on various computational platforms, thus representing a new paradigm approach for protein design. Published by the American Physical Society 2024

Deep generative design of RNA aptamers using structural predictions.

RNAs represent a class of programmable biomolecules capable of performing diverse biological functions. Recent studies have developed accurate RNA three-dimensional structure prediction methods, which may enable new RNAs to be designed in a structure-guided manner. Here, we develop a structure-to-sequence deep learning platform for the de novo generative design of RNA aptamers. We show that our approach can design RNA aptamers that are predicted to be structurally similar, yet sequence dissimilar, to known light-up aptamers that fluoresce in the presence of small molecules. We experimentally validate several generated RNA aptamers to have fluorescent activity, show that these aptamers can be optimized for activity in silico, and find that they exhibit a mechanism of fluorescence similar to that of known light-up aptamers. Our results demonstrate how structural predictions can guide the targeted and resource-efficient design of new RNA sequences.

Recalibration of MARTINI-3 Parameters for Improved Interactions between Peripheral Proteins and Lipid Bilayers.

The MARTINI force field is one of the most used coarse-grained models for biomolecular simulations. Many limitations of the model including the protein-protein overaggregation have been improved in its latest version, MARTINI-3. In this study, we investigate the efficacy of the MARTINI-3 parameters for capturing the interactions of peripheral proteins with model plasma membranes. Particularly, we consider two classes of proteins, namely, annexin and epsin, which are known to generate negative and positive membrane curvatures, respectively. We find that current MARTINI-3 parameters are not able to correctly describe the protein-membrane interface and the protein-induced membrane curvatures for any of these proteins. The problem arises due to the lack of proper hydrophobic interactions between the protein residues and lipid tails. Making systematic adjustments, we show that a combination of reduction in the protein-water interactions and enhancement of protein-lipid hydrophobic interactions is essential for accurate prediction of the interfacial structure including the protein-induced membrane curvature. Next, we apply our model to a couple of other peripheral proteins, namely, Snf7, a core component of the ESCRT-III complex, and the PH domain of evectin-2. We find that our model captures the protein-membrane interfacial structure much more accurately than the MARTINI-3 model for all of the peripheral proteins considered in this study. However, the strategy described in this study may not be suitable for oligomeric transmembrane proteins where protein-protein hydrophobic interactions should be increased instead of protein-lipid hydrophobic interactions.

Using the RNAstructure Software Package to Predict Conserved RNA Structures.

The structures of many non-coding RNAs (ncRNA) are conserved by evolution to a greater extent than their sequences. By predicting the conserved structure of two or more homologous sequences, the accuracy of secondary structure prediction can be improved as compared to structure prediction for a single sequence. Here, we provide protocols for the use of four programs in the RNAstructure suite topredict conserved structures: Multilign, TurboFold, Dynalign, and PARTS. TurboFold iteratively aligns multiple homologous sequences and estimates the pairing probabilities for the conserved structure. Dynalign, PARTS, and Multilign are dynamic programming algorithms that simultaneously align sequences and identify the common secondary structure. Dynalign uses a pair of homologs and finds the lowest free energy common structure. PARTS uses a pair of homologs and estimates pairing probabilities from the base pairing probabilities estimated for each sequence. Multilign uses two or morehomologs and finds the lowest free energy common structure using multiple pairwise calculations with Dynalign. It scales linearly with the number of sequences. We outline the strengths of each program. These programs can be run through web servers, on the command line, or with graphical user interfaces. © 2024 Wiley Periodicals LLC. Basic Protocol 1: Predicting a structure conserved in three or more sequences with the RNAstructure web server Basic Protocol 2: Predicting a structure conserved in two sequences with the RNAstructure web server Alternative Protocol 1: Predicting a structure conserved in multiple sequences in the RNAstructure graphical user interface Alternative Protocol 2: Predicting a structure conserved in two sequences with Dynalign in the RNAstructure graphical user interface Alternative Protocol 3: Running TurboFold on the command line.

Crystal Structure Prediction Using Generative Adversarial Network with Data-Driven Latent Space Fusion Strategy.

Crystal structure prediction (CSP) is an important field of material design. Herein, we propose a novel generative adversarial network model, guided by a data-driven approach and incorporating the real physical structure of crystals, to address the complexity of high-dimensional data and improve prediction accuracy in materials science. The model, termed GAN-DDLSF, introduces a novel sampling method called data-driven latent space fusion (DDLSF), which aims to optimize the latent space of generative adversarial networks (GANs) by combining the statistical properties of real data with a standard Gaussian distribution, effectively mitigating the "mode collapse" problem prevalent in GANs. Our approach introduces a more refined generation mechanism specifically for binary crystal structures such as gallium nitride (GaN). By optimizing for the specific crystallographic features of GaN while maintaining structural rationality, we achieve higher precision and efficiency in predicting and designing structures for this particular material system. The model generates 9321 GaN binary crystal structures, with 16.59% reaching a stable state and 24.21% found to be metastable. These results can significantly enhance the accuracy of crystal structure predictions and provide valuable insights into the potential of the GAN-DDLSF approach for the discovery and design of binary, ternary, and multinary materials, offering new perspectives and methods for materials science research and applications.

Elucidating Protein Structures in the Gas Phase: Traversing Configuration Space with Biasing Methods.

Achieving accurate characterization of protein structures in the gas phase continues to be a formidable challenge. To tackle this issue, the present study employs Molecular Dynamics (MD) simulations in tandem with enhanced sampling techniques (methods designed to efficiently explore protein conformations). The objective is to identify suitable structures of proteins by contrasting their calculated Collision Cross-Section (CCS) with those observed experimentally. Significant discrepancies were observed between the initial MD-simulated and experimentally measured CCS values through Ion Mobility-Mass Spectrometry (IMS-MS). To bridge this gap, we employed two distinct enhanced sampling methods, Harmonic Biasing Potential and Adaptive Biasing Force, which help the proteins overcome energy barriers to adopt more compact configurations. These techniques leverage the radius of gyration as a reaction coordinate (guiding parameter), guiding the system toward compressed states that potentially match experimental configurations more closely. The guiding forces are only employed to overcome existing barriers and are removed to allow the protein to naturally arrive at a potential gas phase configuration. The results demonstrated close alignment (within ∼4%) between simulated and experimental CCS values despite using different strengths and/or methods, validating their efficacy. This work lays the groundwork for future studies aimed at optimizing biasing methods and expanding the collective variables used for more accurate gas-phase structural predictions.

Easy and accurate protein structure prediction using ColabFold.

Since its public release in 2021, AlphaFold2 (AF2) has made investigating biological questions, by using predicted protein structures of single monomers or full complexes, a common practice. ColabFold-AF2 is an open-source Jupyter Notebook inside Google Colaboratory and a command-line tool that makes it easy to use AF2 while exposing its advanced options. ColabFold-AF2 shortens turnaround times of experiments because of its optimized usage of AF2's models. In this protocol, we guide the reader through ColabFold best practices by using three scenarios: (i) monomer prediction, (ii) complex prediction and (iii) conformation sampling. The first two scenarios cover classic static structure prediction and are demonstrated on the human glycosylphosphatidylinositol transamidase protein. The third scenario demonstrates an alternative use case of the AF2 models by predicting two conformations of the human alanine serine transporter 2. Users can run the protocol without computational expertise via Google Colaboratory or in a command-line environment for advanced users. Using Google Colaboratory, it takes <2 h to run each procedure. The data and code for this protocol are available at https://protocol.colabfold.com .

The success rate of processed predicted models in molecular replacement: implications for experimental phasing in the AlphaFold era.

The availability of highly accurate protein structure predictions from AlphaFold2 (AF2) and similar tools has hugely expanded the applicability of molecular replacement (MR) for crystal structure solution. Many structures can be solved routinely using raw models, structures processed to remove unreliable parts or models split into distinct structural units. There is therefore an open question around how many and which cases still require experimental phasing methods such as single-wavelength anomalous diffraction (SAD). Here, this question is addressed using a large set of PDB depositions that were solved by SAD. A large majority (87%) could be solved using unedited or minimally edited AF2 predictions. A further 18 (4%) yield straightforwardly to MR after splitting of the AF2 prediction using Slice'N'Dice, although different splitting methods succeeded on slightly different sets of cases. It is also found that further unique targets can be solved by alternative modelling approaches such as ESMFold (four cases), alternative MR approaches such as ARCIMBOLDO and AMPLE (two cases each), and multimeric model building with AlphaFold-Multimer or UniFold (three cases). Ultimately, only 12 cases, or 3% of the SAD-phased set, did not yield to any form of MR tested here, offering valuable hints as to the number and the characteristics of cases where experimental phasing remains essential for macromolecular structure solution.

Towards the accurate modelling of antibody-antigen complexes from sequence using machine learning and information-driven docking.

Antibody-antigen complex modelling is an important step in computational workflows for therapeutic antibody design. While experimentally determined structures of both antibody and the cognate antigen are often not available, recent advances in machine learning-driven protein modelling have enabled accurate prediction of both antibody and antigen structures. Here, we analyse the ability of protein-protein docking tools to use machine learning generated input structures for information-driven docking. In an information-driven scenario, we find that HADDOCK can generate accurate models of antibody-antigen complexes using an ensemble of antibody structures generated by machine learning tools and AlphaFold2 predicted antigen structures. Targeted docking using knowledge of the complementary determining regions on the antibody and some information about the targeted epitope allows the generation of high-quality models of the complex with reduced sampling, resulting in a computationally cheap protocol that outperforms the ZDOCK baseline. The source code of HADDOCK3 is freely available at github.com/haddocking/haddock3. The code to generate and analyse the data is available at github.com/haddocking/ai-antibodies. The full runs, including docking models from all modules of a workflow have been deposited in our lab collection (data.sbgrid.org/labs/32/1139) at the SBGRID data repository.