An easy-to-use three-dimensional protein-structure-prediction online platform "DPL3D" based on deep learning algorithms.

Structural genomics provides an important approach for characterizing and understanding systems biology. As a step toward better integrating protein three-dimensional (3D) structural information in cancer systems biology, we have constructed a Human Cancer Pathway Protein Interaction Network (HCPIN) by analysis of several classical cancer-associated signaling pathways and their physical protein-protein interactions. Many well known cancer-associated proteins play central roles as "hubs" or "bottlenecks" in the HCPIN. At least half of HCPIN proteins are either directly associated with or interact with multiple signaling pathways. Although some 45% of residues in these proteins are in sequence segments that meet criteria sufficient for approximate homology modeling (Basic Local Alignment Search Tool (BLAST) E-value <10(-6)), only approximately 20% of residues in these proteins are structurally covered using high accuracy homology modeling criteria (i.e. BLAST E-value <10(-6) and at least 80% sequence identity) or by actual experimental structures. The HCPIN Website provides a comprehensive description of this biomedically important multipathway network together with experimental and homology models of HCPIN proteins useful for cancer biology research. To complement and enrich cancer systems biology, the Northeast Structural Genomics Consortium is targeting >1000 human proteins and protein domains from the HCPIN for sample production and 3D structure determination. The long range goal of this effort is to provide a comprehensive 3D structure-function database for human cancer-associated proteins and protein complexes in the context of their interaction networks. The network-based target selection (BioNet) approach described here is an example of a general strategy for targeting co-functioning proteins by structural genomics projects.

Recent advancements in artificial intelligence (AI) have accelerated the prediction of unknown protein structures. However, accurately predicting the three-dimensional (3D) structures of fusion proteins remains a difficult task because the current AI-based protein structure predictions are focused on the WT proteins rather than on the newly fused proteins in nature. Following the central dogma of biology, fusion proteins are translated from fusion transcripts, which are made by transcribing the fusion genes between two different loci through the chromosomal rearrangements in cancer. Accurately predicting the 3D structures of fusion proteins is important for understanding the functional roles and mechanisms of action of new chimeric proteins. However, predicting their 3D structure using a template-based model is challenging because known template structures are often unavailable in databases. Deep learning (DL) models that utilize multi-level protein information have revolutionized the prediction of protein 3D structures. In this review paper, we highlighted the latest advancements and ongoing challenges in predicting the 3D structure of fusion proteins using DL models. We aim to explore both the advantages and challenges of employing AlphaFold2, RoseTTAFold, tr-Rosetta and D-I-TASSER for modelling the 3D structures. HIGHLIGHTS: This review provides the overall pipeline and landscape of the prediction of the 3D structure of fusion protein. This review provides the factors that should be considered in predicting the 3D structures of fusion proteins using AI approaches in each step. This review highlights the latest advancements and ongoing challenges in predicting the 3D structure of fusion proteins using deep learning models. This review explores the advantages and challenges of employing AlphaFold2, RoseTTAFold, tr-Rosetta, and D-I-TASSER to model 3D structures.

While of primary importance in both the biomedical and therapeutic fields, peptides suffer from a relative lack of dedicated tools to predict efficiently and accurately their 3D structures despite being a crucial step in understanding their physio-pathological function or designing new drugs. In recent years, deep-learning methods have enabled a major breakthrough for the protein 3D structure prediction approaches, allowing to predict protein 3D structures with a near-experimental accuracy for nearly any protein sequence. This present study aims at confronting some of these new methods (AlphaFold2, RoseTTAFold2, and ESMFold) for the peptides' 3D structure prediction problem and evaluating their performance. All methods produced high-quality results, but their overall performance is lower as compared to the prediction of protein 3D structures. We also identified a few structural features that impede the ability to produce high-quality peptide structure predictions. These findings point out the discrepancy that still exists between the protein and peptide 3D structure prediction methods and underline a few cases where the generated peptide structures should be used very cautiously.

Predicting protein-protein interaction with interpretable bilinear attention network.

Databases are the computerized platform where information is stored and can be retrieved easily by public users. Biological databases are the repositories of biological data. These biological data libraries contain facts and figures related to various disciplines of research including genomics, proteomics, microarray technology, metabolomics and phylogenetics. By using biological databases, a broad collection of essential biological information can be exploited ranging from function, structure and localization of gene, clinical consequences of mutation to similarity index among biological sequences and structures. Nowadays, different kinds of biological databases are available on the web. The present write up focuses on biological databases and bioinformatics tools for protein structure analysis. This review also aims to elaborate the searching schemes, available in different structural databases. The wide variety of different levels and types of information content related to 3D protein structures are available on web-based databases. Regarding the biological functions and 3D structures of various proteins, these databases provide a huge range of useful links, schematic diagrams as well as strategies for detailed analysis of proteins and other macromolecules structures. 3D structural illustration of proteins stored in structural databases is determined and visualized by X-ray crystallography, electron microscopy and NMR spectroscopy. On regular basis, a large number of protein structures are submitted by structural biologists, updated and curated by subject experts. Most familiar biological databases that store 3D protein and other macromolecules structures include, PDB, 3D Genomics, CATH, &amp; SCOP. These databases contain valuable information of overall protein structures, domains and motif structures, protein-protein complex systems and complex of protein with other biomolecules.

Cluster Quality based Non-Reductional (CQNR) oversampling technique and effector protein predictor based on 3D structure (EPP3D) of proteins

Objective: Severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) became one of the most important health problems of the 21st century. Non-structural protein-13 (nsp13/helicase) plays an important role in the replication of the viral genome and the viral life cycle. The SARS-CoV-2 genome has undergone thousands of mutations since the disease first appeared. Mutations pose a threat to the validity of therapeutics due to changes in protein structure. Modeling alterations caused by mutations in the viral proteome contributes to the development of effective antivirals. The changes in protein structure and stability caused by mutations seen in European isolates of SARS-CoV-2 were analyzed in the study with the aim of contributing to studies on the development of new anti-virals and the validity of existing therapeutics. Methods: The changes in protein structure after mutation were modeled with deep learning algorithms. The alterations in protein stability were analyzed by SDM2, mCSM, DUET and DynaMut2. Results: The mutation analysis revealed four (Pro77Leu, Gly170Ser, Tyr324Cys, and Arg392Cys) missense mutations in the nsp13 protein in European isolates of SARS-CoV-2. Mutations caused changes in protein structure (rmsd 0.294 Å) and stability (-.58 ≤ ΔΔG ≤ .003 kcal.mol-1). The atomic interactions formed by the mutant residues in the three-dimensional conformation of the protein have changed. Conclusion: The mutations seen in European isolates for nsp13 of SARS-CoV-2 may lead to the emergence of different phenotypes in terms of viral activity. For this reason, the study may contribute to the success of the fight against the virus with different treatment approaches in different regions.

Although the number of known protein structures is increasing, the number of protein sequences without determined structures is still much larger. Three-dimensional (3D) protein structure information helps in the understanding of functional mechanisms, but solving structures by X-ray crystallography or NMR is often a lengthy and difficult process. A relatively fast way of determining a protein's 3D structure is to construct a computer model using homologous sequence and structure information. Much work has gone into algorithms that comprise the ORCHESTRAR homology modeling program in the SYBYL software package. This novel homology modeling tool combines algorithms for modeling conserved cores, variable regions, and side chains. The paradigm of using existing knowledge from multiple templates and the underlying protein environment knowledgebase is used in all of these algorithms, and will become even more powerful as the number of experimentally derived protein structures increases. To determine how ORCHESTRAR compares to Composer (a broadly used, but an older tool), homology models of 18 proteins were constructed using each program so that a detailed comparison of each step in the modeling process could be carried out. Proteins modeled include kinases, dihydrofolate reductase, HIV protease, and factor Xa. In almost all cases ORCHESTRAR produces models with lower root-mean-squared deviation (RMSD) values when compared with structures determined by X-ray crystallography or NMR. Moreover, ORCHESTRAR produced a homology model for three target sequences where Composer failed to produce any. Data for RMSD comparisons between structurally conserved cores, structurally variable regions, side-chain conformations are presented, as well as analyses of active site and protein-protein interface configurations.

Pharmacogenomics is the study of the genetic basis for individual variation in response to drugs and other xenobiotics. Successful prediction of effects of genetic variations that change encoded amino acid sequences on protein function and their consequent biomedical implications depends on three-dimensional (3D) structures of the encoded amino acid sequences. To bridge sequence to function, thus facilitating an in-depth pharmacogenomic study, we tested the feasibility of the use of a semi-computational approach to predict 3D structures of rabbit and human indolethylamine N-methyltransferases (INMTs) from their amino acid sequences, which share less than 26% sequence identity with known protein 3D structures. Herein, we report 3D models of INMTs predicted by using the crystal structure of rat catechol O-methyltransferase as a template, testing of the models both computationally and experimentally, and successful use of the models in retrospective prediction of the effects of genetic polymorphisms and in identification of residues that contribute to observed species-specific differences in substrate affinity. The results encourage the use of the semi-computational approach to predict 3D protein structures for use in pharmacogenomic studies when de novo prediction of protein 3D structures from their amino acid sequences is still not feasible and X-ray crystallography and/or solution nuclear magnetic resonance spectroscopy can only determine 3D structures for a small number of known amino acid sequences.

Artificial intelligence (AI) has transformed the landscape of protein structural and functional prediction, significantly advancing the accuracy and efficiency of these processes. Currently, AI-driven methods, especially deep learning algorithms, enable the prediction of protein 3D structures from amino acid sequences with unprecedented precision. Artificial intelligence (AI) has emerged as a transformative force in the field of protein science, offering powerful tools for the structural and functional prediction of proteins. AI models use vast databases of known protein structures and leverage evolutionary information from multiple sequence alignments or protein language models to infer spatial conformations of proteins. Deep neural networks, convolutional neural networks, and graph-based models enhance prediction accuracy beyond traditional homology or ab initio methods. AlphaFold2’s breakthrough in CASP14 demonstrated near-experimental accuracy for many proteins, ushering in a new era of AI-based structural biology. AI-driven protein structure and function prediction tools are democratizing access to complex biological data, making it possible for many research groups to accelerate discovery without expensive and time-consuming experiments. Machine learning models, such as DeepGO-SE, utilize pretrained protein language models alongside biological knowledge and protein interaction networks to predict Gene Ontology functions. These models improve prediction accuracy even for proteins with unknown interactions. This review discusses the latest advancements in AI-driven methodologies, including deep learning models and large language models, highlighting their significant contributions to resolving protein structures, functional annotation, and interaction mapping. The article summarizes current achievements, evaluates the strengths and limitations of AI approaches, and outlines future prospects for integrating AI with experimental data to accelerate discoveries in proteomics and drug discovery.

Information of 3D protein structure plays an important role in various fields, including health, biotechnology, biomaterials, and so on. Knowledge of the 3D structure of proteins will help in understanding the interactions that can take place with other molecules such as drug molecules and other effector molecules. To get the 3D protein structure, it is generally carried out experimentally using x-ray diffraction and NMR (Nuclear Magnetic Resonance) methods. The experimental method requires a relatively long time and expertise to handle the completion of the structure. Until now, not all protein structures can be determined because the level of complexity varies from one protein to another. One approach is to use machine learning that leverages evolution information and deep learning method. The results given can improve the accuracy of the prediction compared to the conventional approach. However, the level of accuracy is still influenced by the number of homologous proteins in the database. Therefore, this study propose to replace the process of extracting the evolutionary information from Multiple Sequence Analysis (MSA) into Transformer-based self-supervised pre-trained. To test these changes, an experiment was carried out on a 3D protein structure prediction model based on Long Short-Term Memory (LSTM) and Universal Transformer. The proposed method results show a decrease in the value of distance Root-Mean Square Deviation (dRMSD) of 0.561 Angstrom and Root Mean Square Error (RMSE) torsion angle of 0.11 degree on the Universal Transformer predictor. The results of the T-Test show that the decrease in the two indicators shows a significant result. Therefore, pre-trained data can be used as an evolutionary information only for the Universal Transformer predictor.

In recent years, various computational methods have been developed to predict the three-dimensional (3D) structures of RNAs. Due to its hierarchical folding property, RNA secondary (2D) structure is usually used as input for 3D structure prediction to improve accuracy and efficiency. However, the extent to which the accuracy of input 2D structure affects the performance of 3D structure prediction remains to be further investigated. Additionally, whether and how the input base-pairing interactions are modified during the 3D structure modeling process is another question worth exploring. To address these issues, here we comprehensively benchmark five representative 3D structure prediction models on extensive datasets, using 2D structures of varied accuracies as input. Our results indicate that there is a pervasive cross-talk between RNA 2D and 3D structure predictions, where the performance dependence of 3D structure prediction on the accuracy of input 2D structure is closely associated with the 3D model’s ability to modify the input base-pairing interactions during structure modeling. Furthermore, we also observed that RNA 3D structure prediction performance is more sensitive to the occurrence of false positive base pairs in the input 2D structure than to true positive base pairs, suggesting a worthy direction to further improve the model performance.Author summaryThree-dimensional (3D) structural modeling of RNAs with large sizes and complex topologies remains challenging despite the availability of (predicted) 2D structures as constraints. Exploring the potential cross-talk between 2D and 3D structure predictions is a worthy direction to improve the performance of RNA structure modeling. In this study, we found that all tested popular RNA 3D structure prediction models were able to modify the original base pairing interactions contained in the input 2D structure during the 3D modeling process. Especially, all of these models presented increased F1-score for the optimal combination of RNA 2D and 3D structure prediction models. The results suggest that a worthwhile direction to further improve the performance of RNA 3D structure prediction is to minimize the incidence of incorrectly predicted base pairing interactions during modeling process without compromising or even improving the presence of correct interactions in the input 2D structure.

Structural Features can be Unconserved in Proteins with Similar Folds: An Analysis of Side-chain to Side-chain Contacts Secondary Structure and Accessibility

The p7 protein of Hepatitis C virus (HCV) is a small integral membrane protein consists of 63 amino acids that is crucial for assembly and release of infectious virions. The p7 protein forms viral ion channels that can change membrane permeability, and could thus be considered as an antiviral target for drug design.The present study was designed to carry out a phylogenetic analysis and secondary and 3-D structure prediction of the p7 protein in the HCV genotype 4a detected in different countries. This is to examine the diversity of this genotype in different geographical regions and determine conserved regions that can be considered as a potential antiviral target for drug design. Molecular evolutionary and phylogenetic analysis using Jalview program showed that the HCV p7 gene of the genotype 4a isolates detected in Egypt was closely related to their counterpart in Germany. The phylogenetic analysis of HCV p7 genotype isolates from several parts of the world showing high genomic diversity of genotype 4 where the HCV-p7 protein genotype 4a dendrogram illustrates that the Egyptian isolates were classified into four clusters. Secondary structure predictions using JNetpred algorithm suggested that the p7 protein of the HCV genotype 4 contain one alpha Helix at 20-34(HPRLVRHLLHLHC amino acids) and three β-sheets at 5-7(GSVamino acids), 13-16(QCCF amino acids) and 43-47(CCYLR amino acids. The 3-D structure prediction model for the p7 protein of different HCV genotype 4a isolates (using J.mol) showed two coiled-coils. The α-helical coiled coil is a principal subunit oligomerization motifs in approximately 10% of all protein sequences.

The prediction and analysis of the three- dimensional (3D) structure of proteins is a key research problem in Structural Bioinformatics. The 1990's Genome Projects resulted in a large increase in the number of available protein sequences. However, the number of identified 3D protein structures have not followed the same growth trend. Currently, the number of available protein sequences greatly exceeds the number of known 3D structures. Many computational methodologies, systems and algorithms have been proposed to address the protein structure prediction problem. However, the problem still remains challenging because of the complexity and high dimensionality of a protein conformational search space. The most significant progress in the last Critical Assessment of protein Structure Prediction was achieved by methods that use database information. Nevertheless, a major challenge remains in the development of better strategies for template identification and representation. This article describes a computational strategy to acquire and represent structural information of experimentally determined 3D protein structures. A clustering strategy was combined with artificial neural networks in order to extract structural information from experimental protein structure templates. In the proposed strategy, the main efforts focus on the acquisition of useful and accurate structural information from 3D protein templates stored in the Protein Data Bank (PDB). The proposed method was tested in twenty protein sequences whose sizes vary from 14 to 70 amino acid residues. Our results show that the proposed method is a good way to extract and represent valuable information obtained from the PDB and also significantly reduce the 3D protein conformational search space.