Proteins Of Unknown Structure Research Articles

Artificial intelligence in fusion protein three-dimensional structure prediction: Review and perspective.

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.

Bioinspired Algorithms for Multiple Sequence Alignment: A Systematic Review and Roadmap

Multiple Sequence Alignment (MSA) plays a pivotal role in bioinformatics, facilitating various critical biological analyses, including the prediction of unknown protein structures and functions. While numerous methods are available for MSA, bioinspired algorithms stand out for their efficiency. Despite the growing research interest in addressing the MSA challenge, only a handful of comprehensive reviews have been undertaken in this domain. To bridge this gap, this study conducts a thorough analysis of bioinspired-based methods for MSA through a systematic literature review (SLR). By focusing on publications from 2010 to 2024, we aim to offer the most current insights into this field. Through rigorous eligibility criteria and quality standards, we identified 45 relevant papers for review. Our analysis predominantly concentrates on bioinspired-based techniques within the context of MSA. Notably, our findings highlight Genetic Algorithm and Memetic Optimization as the most commonly utilized algorithms for MSA. Furthermore, benchmark datasets such as BAliBASE and SABmark are frequently employed in evaluating MSA solutions. Structural-based methods emerge as the preferred approach for assessing MSA solutions, as revealed by our systematic literature review. Additionally, this study explores current trends, challenges, and unresolved issues in the realm of bioinspired algorithms for MSA, offering practitioners and researchers valuable insights and comprehensive understanding of the field.

A Perspective on the Prospective Use of AI in Protein Structure Prediction.

AlphaFold2 (AF2) and RoseTTaFold (RF) have revolutionized structural biology, serving as highly reliable and effective methods for predicting protein structures. This article explores their impact and limitations, focusing on their integration into experimental pipelines and their application in diverse protein classes, including membrane proteins, intrinsically disordered proteins (IDPs), and oligomers. In experimental pipelines, AF2 models help X-ray crystallography in resolving the phase problem, while complementarity with mass spectrometry and NMR data enhances structure determination and protein flexibility prediction. Predicting the structure of membrane proteins remains challenging for both AF2 and RF due to difficulties in capturing conformational ensembles and interactions with the membrane. Improvements in incorporating membrane-specific features and predicting the structural effect of mutations are crucial. For intrinsically disordered proteins, AF2's confidence score (pLDDT) serves as a competitive disorder predictor, but integrative approaches including molecular dynamics (MD) simulations or hydrophobic cluster analyses are advocated for accurate dynamics representation. AF2 and RF show promising results for oligomeric models, outperforming traditional docking methods, with AlphaFold-Multimer showing improved performance. However, some caveats remain in particular for membrane proteins. Real-life examples demonstrate AF2's predictive capabilities in unknown protein structures, but models should be evaluated for their agreement with experimental data. Furthermore, AF2 models can be used complementarily with MD simulations. In this Perspective, we propose a "wish list" for improving deep-learning-based protein folding prediction models, including using experimental data as constraints and modifying models with binding partners or post-translational modifications. Additionally, a meta-tool for ranking and suggesting composite models is suggested, driving future advancements in this rapidly evolving field.

Molecular modeling, simulation and docking of Rv1250 protein from Mycobacterium tuberculosis.

Computational prediction and protein structure modeling have come to the aid of various biological problems in determining the structure of proteins. These technologies have revolutionized the biological world of research, allowing scientists and researchers to gain insights into their biological questions and design experimental research much more efficiently. Pathogenic Mycobacterium spp. is known to stay alive within the macrophages of its host. Mycobacterium tuberculosis is an acid-fast bacterium that is the most common cause of tuberculosis and is considered to be the main cause of resistance of tuberculosis as a leading health issue. The genome of Mycobacterium tuberculosis contains more than 4,000 genes, of which the majority are of unknown function. An attempt has been made to computationally model and dock one of its proteins, Rv1250 (MTV006.22), which is considered as an apparent drug-transporter, integral membrane protein, and member of major facilitator superfamily (MFS). The most widely used techniques, i.e., homology modeling, molecular docking, and molecular dynamics (MD) simulation in the field of structural bioinformatics, have been used in the present work to study the behavior of Rv1250 protein from M. tuberculosis. The structure of unknown TB protein, i.e., Rv1250 was retrived using homology modeling with the help of I-TASSER server. Further, one of the sites responsible for infection was identified and docking was done by using the specific Isoniazid ligand which is an inhibitor of this protein. Finally, the stability of protein model and analysis of stable and static interaction between protein and ligand molecular dynamic simulation was performed at 100ns The designing of novel Rv1250 enzyme inhibitors is likely achievable with the use of proposed predicted model, which could be helpful in preventing the pathogenesis caused by M. tuberculosis. Finally, the MD simulation was done to evaluate the stability of the ligand for the specific protein.

Application of deep learning methods for protein structure prediction

The rapid development of machine learning and deep learning methods has led to their widespread use in many areas of research, including their use in solving problems of proteomics. Today, machine learning methods play an important role in predicting the three-dimensional structural components of proteins by interpreting how protein sequences and their homology govern interresidual contacts and structural organization. Recent CASPs (Critical Assessment of Structure Prediction) assessing the state of the art in modeling protein structure based on amino acid sequence have demonstrated significant progress in modeling structures without the use of structural templates (historically “ab initio” modeling). The progress was driven by the successful application of deep learning techniques to predict the distances between residuals. In turn, these results led to a significant increase in the accuracy of the three-dimensional structure, provided that a sufficient number of sequences are known for a family of proteins. In addition, the number of sequences required for alignment has been significantly reduced, and the accuracy of templatebased models has also improved significantly. This paper provides an overview of recent advances in the application of deep learning methods used to predict the three-dimensional structure of a protein. The possibilities of using neural networks to identify unknown protein structures and functions of proteins, which is one of the most important tasks of proteomics, are also considered. Problems that have yet to be solved are described, but it is expected that in the near future the described methods will play a decisive role in the structural bioinformatics of proteins.

Validated determination of NRG1 Ig-like domain structure by mass spectrometry coupled with computational modeling

High resolution hydroxyl radical protein footprinting (HR-HRPF) is a mass spectrometry-based method that measures the solvent exposure of multiple amino acids in a single experiment, offering constraints for experimentally informed computational modeling. HR-HRPF-based modeling has previously been used to accurately model the structure of proteins of known structure, but the technique has never been used to determine the structure of a protein of unknown structure. Here, we present the use of HR-HRPF-based modeling to determine the structure of the Ig-like domain of NRG1, a protein with no close homolog of known structure. Independent determination of the protein structure by both HR-HRPF-based modeling and heteronuclear NMR was carried out, with results compared only after both processes were complete. The HR-HRPF-based model was highly similar to the lowest energy NMR model, with a backbone RMSD of 1.6 Å. To our knowledge, this is the first use of HR-HRPF-based modeling to determine a previously uncharacterized protein structure.

Recent Applications of Deep Learning Methods on Evolution- and Contact-Based Protein Structure Prediction.

The new advances in deep learning methods have influenced many aspects of scientific research, including the study of the protein system. The prediction of proteins’ 3D structural components is now heavily dependent on machine learning techniques that interpret how protein sequences and their homology govern the inter-residue contacts and structural organization. Especially, methods employing deep neural networks have had a significant impact on recent CASP13 and CASP14 competition. Here, we explore the recent applications of deep learning methods in the protein structure prediction area. We also look at the potential opportunities for deep learning methods to identify unknown protein structures and functions to be discovered and help guide drug–target interactions. Although significant problems still need to be addressed, we expect these techniques in the near future to play crucial roles in protein structural bioinformatics as well as in drug discovery.

DeepMSA: constructing deep multiple sequence alignment to improve contact prediction and fold-recognition for distant-homology proteins.

The success of genome sequencing techniques has resulted in rapid explosion of protein sequences. Collections of multiple homologous sequences can provide critical information to the modeling of structure and function of unknown proteins. There are however no standard and efficient pipeline available for sensitive multiple sequence alignment (MSA) collection. This is particularly challenging when large whole-genome and metagenome databases are involved. We developed DeepMSA, a new open-source method for sensitive MSA construction, which has homologous sequences and alignments created from multi-sources of whole-genome and metagenome databases through complementary hidden Markov model algorithms. The practical usefulness of the pipeline was examined in three large-scale benchmark experiments based on 614 non-redundant proteins. First, DeepMSA was utilized to generate MSAs for residue-level contact prediction by six coevolution and deep learning-based programs, which resulted in an accuracy increase in long-range contacts by up to 24.4% compared to the default programs. Next, multiple threading programs are performed for homologous structure identification, where the average TM-score of the template alignments has over 7.5% increases with the use of the new DeepMSA profiles. Finally, DeepMSA was used for secondary structure prediction and resulted in statistically significant improvements in the Q3 accuracy. It is noted that all these improvements were achieved without re-training the parameters and neural-network models, demonstrating the robustness and general usefulness of the DeepMSA in protein structural bioinformatics applications, especially for targets without homologous templates in the PDB library. https://zhanglab.ccmb.med.umich.edu/DeepMSA/. Supplementary data are available at Bioinformatics online.

Fourier-Transform Infrared (FTIR) Spectroscopy Transmission Technique for Analysis of Characteristic Infrared Bands of Peptide Linkage using Dried Proteins

Protein activity is well dependent on its conformational structure and thus secondary structure prediction and determination of proteins are vital requirements. FTIR is rapidly coming up for protein structural studies, as an alternative to circular dichroism. However, FTIR spectra are primarily recorded using protein samples prepared in water, organic solvents, deuterium oxide, and membranes. FTIR spectra are thus dominated by the solvent spectrum and subtraction of the solvent spectrum introduces severe spectral distortions. In aqueous protein solutions, a strong water signal overlaps amide I bands (C=O stretching; 1700-1600 cm−1), thus masking most of the protein secondary structure information. Further, most of the purified proteins including protein drug formulations are more stable in the solid state than in the liquid state. Hence there is a need to explore the use of FTIR for various physical states of proteins. In the present study, we showed that protein secondary structure can be analyzed in a dried state; using potassium bromide (KBr) pellet method for dried known proteins (bovine serum albumin, proteinase K, fibrinogen etc). KBr pellet method gave excellent spectra for all the used dried proteins; shown by the appearance of characteristic infrared bands of peptide linkage (Amide A, B, I-VI) without much interference from water, buffers, stabilizing factors, co-factors etc. We also found that KBr pellet method can analyze a range of protein types; globular to fibrous. Thus KBr method is a great tool for prediction/determination of the secondary structure of unknown proteins and quality control of solid-state protein drug formulations.

An improved protein structure evaluation using a semi-empirically derived structure property

BackgroundIn the backdrop of challenge to obtain a protein structure under the known limitations of both experimental and theoretical techniques, the need of a fast as well as accurate protein structure evaluation method still exists to substantially reduce a huge gap between number of known sequences and structures. Among currently practiced theoretical techniques, homology modelling backed by molecular dynamics based optimization appears to be the most popular one. However it suffers from contradictory indications of different validation parameters generated from a set of protein models which are predicted against a particular target protein. For example, in one model Ramachandran Score may be quite high making it acceptable, whereas, its potential energy may not be very low making it unacceptable and vice versa. Towards resolving this problem, the main objective of this study was fixed as to utilize a simple experimentally derived output, Surface Roughness Index of concerned protein of unknown structure as an intervening agent that could be obtained using ordinary microscopic images of heat denatured aggregates of the same protein.ResultIt was intriguing to observe that direct experimental knowledge of the concerned protein, however simple it may be, might give insight on acceptability of its particular structural model out of a confusion set of models generated from database driven comparative technique for structure prediction. The result obtained from a widely varying structural class of proteins indicated that speed of protein structure evaluation can be further enhanced without compromising with accuracy by recruiting simple experimental output.ConclusionIn this work, a semi-empirical methodological approach was provided for improving protein structure evaluation. It showed that, once structure models of a protein were obtained through homology technique, the problem of selection of a best model out of a confusion set of Pareto-optimal structures could be resolved by employing a structure agent directly obtainable through experiment with the same protein as experimental ingredient. Overall, in the backdrop of getting a reasonably accurate protein structure of pathogens causing epidemics or biological warfare, such approach could be of use as a plausible solution for fast drug design.

Revealing Unknown Protein Structures Using Computational Conformational Sampling Guided by Experimental Hydrogen-Exchange Data.

Both experimental and computational methods are available to gather information about a protein’s conformational space and interpret changes in protein structure. However, experimentally observing and computationally modeling large proteins remain critical challenges for structural biology. Our work aims at addressing these challenges by combining computational and experimental techniques relying on each other to overcome their respective limitations. Indeed, despite its advantages, an experimental technique such as hydrogen-exchange monitoring cannot produce structural models because of its low resolution. Additionally, the computational methods that can generate such models suffer from the curse of dimensionality when applied to large proteins. Adopting a common solution to this issue, we have recently proposed a framework in which our computational method for protein conformational sampling is biased by experimental hydrogen-exchange data. In this paper, we present our latest application of this computational framework: generating an atomic-resolution structural model for an unknown protein state. For that, starting from an available protein structure, we explore the conformational space of this protein, using hydrogen-exchange data on this unknown state as a guide. We have successfully used our computational framework to generate models for three proteins of increasing size, the biggest one undergoing large-scale conformational changes.

Predicting the protein structure using random forest approach

Predicting the secondary structure of proteins is a challenging task. A large variety approaches exist that include observation using equipment’s and theoretical evaluation, in which the optimal structure is determined. The secondary structure determines 3D tertiary structure of protein, on which features and functionalities of protein depend. This paper use classification technique, Random Forest to build a model which is able to determine structure of unknown proteins. The dataset included the amide frequencies of proteins whose structure is known. Machine learning model is developed that can predict the structure of protein that still need to be exploited. The accuracy of the model is determined using ROC curve. The results confirm the performance of the model constructed using amides dataset.

Elucidation of structural and functional characteristics of collagenase from Pseudomonas aeruginosa

Collagenase is specific protease which capable to hydrolyze the native triple helical or denatured collagen. It is a promising enzyme in view of their potential industrial and biomedical applications. Bacterial collagenases may differ significantly owing to their structure and functional characteristics. The process of collagen hydrolysis by collagenase is still unresolved due to imprecise collagenase structural characteristics. Therefore, the present study aimed to understand the molecular, structural, and functional characteristics of collagenase from Pseudomonas aeruginosa using various computational approaches. The physico-chemical features and secondary structure of the collagenase have been explored. The three-dimensional (3D) model of collagenase has been constructed and verified by various tools such as RAMPAGE, verify 3D, and ERRAT. The presence of more than 90% residues in favored region of Ramachandran plot revealed the good quality of protein model. The result of verify3D tool also indicates that 3D structural modeling of the selected collagenase protein is acceptable. The present study provides a detailed overview of the molecular, functional, and structural characteristics of collagenase from P. aeruginosa which might also be helpful to understand the possible structure and functions of unknown proteins.

Big data technique predicts 12% of unknown protein structures

In a landmark achievement for computer-based prediction of protein structure, researchers have reported the three-dimensional conformations of 12% of the protein families that still have unknown structure. The cohort of 600 newly determined protein family structures includes 100 protein folds not found in the Protein Data Bank and 200 membrane proteins (Science 2017, DOI: 10.1126/science.aah4043). A team led by David Baker of the University of Washington modeled the structures of these proteins using its Rosetta algorithm on a distributed network of volunteer computers. Baker’s team initially deposited some of the predicted structures it found into public databases. In the months afterward, structural biologists serendipitously solved—and confirmed—the 3-D conformation of six of these unknown protein families. “The similarity is impressive,” comments Johannes Soding of the Max Planck Institute for Biophysical Chemistry, who was not involved in the work. “And these aren’t simple, small proteins—most have s...

Role of Bioinformatics in the Study of Ionic Channels.

Ionic channels belong to the group of the most important proteins. Not only do they enable transmembrane transport but they are also the key factors for proper cell function. Mutations changing their structure and functionality often lead to severe diseases called channelopathies. On the other hand, transmembrane channels are very difficult objects for experimental studies. Only 2% of experimentally identified structures are transmembrane proteins, while genomic studies show that transmembrane proteins make up 30% of all coded proteins. This gap could be diminished by bioinformatical methods which enable modeling unknown protein structures, functions, transmembrane location, and ligand binding. Several in silico methods dedicated to transmembrane proteins have been developed; some general methods could also be used. They provide the information unavailable from experiments. Current modeling tools use a variety of computational methods, which provide results of surprisingly high quality.

Developing a Multiplexed Quantitative Cross-Linking Mass Spectrometry Platform for Comparative Structural Analysis of Protein Complexes.

Cross-linking mass spectrometry (XL-MS) represents a recently popularized hybrid methodology for defining protein-protein interactions (PPIs) and analyzing structures of large protein assemblies. In particular, XL-MS strategies have been demonstrated to be effective in elucidating molecular details of PPIs at the peptide resolution, providing a complementary set of structural data that can be utilized to refine existing complex structures or direct de novo modeling of unknown protein structures. To study structural and interaction dynamics of protein complexes, quantitative cross-linking mass spectrometry (QXL-MS) strategies based on isotope-labeled cross-linkers have been developed. Although successful, these approaches are mostly limited to pairwise comparisons. In order to establish a robust workflow enabling comparative analysis of multiple cross-linked samples simultaneously, we have developed a multiplexed QXL-MS strategy, namely, QMIX (Quantitation of Multiplexed, Isobaric-labeled cross (X)-linked peptides) by integrating MS-cleavable cross-linkers with isobaric labeling reagents. This study has established a new analytical platform for quantitative analysis of cross-linked peptides, which can be directly applied for multiplexed comparisons of the conformational dynamics of protein complexes and PPIs at the proteome scale in future studies.

A hydrophobic proclivity index for protein alignments

Sequence alignment algorithms are fundamental to modern bioinformatics. Sequence alignments are widely used in diverse applications such as phylogenetic analysis, database searches for related sequences to aid identification of unknown protein domain structures and classification of proteins and protein domains. Additionally, alignment algorithms are integral to the location of related proteins to secure understanding of unknown protein functions, to suggest the folded structure of proteins of unknown structure from location of homologous proteins and/or by locating homologous domains of known 3D structure. For proteins, alignment algorithms depend on information about amino acid substitutions that allows for matching sequences that are similar, but not exact. When primary sequence percent identity falls below about 25%, algorithms often fail to identify proteins that may have similar 3D structure. We have created a hydrophobicity scale and a matching dynamic programming algorithm called TMATCH (unpublished report) that is able to match proteins with remote homologs with similar secondary/tertiary structure, even with very low primary sequence matches. In this paper, we describe how we arrived at the hydrophobic scale, how it provides much more information than percent identity matches and some of the implications for better alignments and understanding protein structure.

Parkinson's disease: Crystals of a toxic core.

An ultra-high-resolution structure of the core segment of assembled α-synuclein — the protein that aggregates in the brains of patients with Parkinson's disease — has been determined. A neurobiologist and a structural biologist discuss the implications of this advance. See Article p.486 The small protein α-synuclein (α-syn) relies on a short segment called NACore (residues 68–78) to form amyloid aggregates responsible for cytotoxicity in Parkinson's disease, but structural information about it has been lacking. Using the frontier-method of micro-electron diffraction (MicroED), David Eisenberg and collaborators have now determined the structure of NACore nanocrystals too small to even be visible by light microscopy. They observe a structure similar to that of toxic fibrils formed by full-length α-syn. The 1.4A resolution of this structure is the highest achieved by any cryo-EM method to date, and the work shows, for the first time, that MicroED can be used to solve previously unknown protein structures.

Structure of the toxic core of α-synuclein from invisible crystals.

The protein α-synuclein is the main component of Lewy bodies, the neuron-associated aggregates seen in Parkinson disease and other neurodegenerative pathologies. An 11-residue segment, which we term NACore, appears to be responsible for amyloid formation and cytotoxicity of human α-synuclein. Here we describe crystals of NACore that have dimensions smaller than the wavelength of visible light and thus are invisible by optical microscopy. As the crystals are thousands of times too small for structure determination by synchrotron X-ray diffraction, we use micro-electron diffraction to determine the structure at atomic resolution. The 1.4 Å resolution structure demonstrates that this method can determine previously unknown protein structures and here yields, to our knowledge, the highest resolution achieved by any cryo-electron microscopy method to date. The structure exhibits protofibrils built of pairs of face-to-face β-sheets. X-ray fibre diffraction patterns show the similarity of NACore to toxic fibrils of full-length α-synuclein. The NACore structure, together with that of a second segment, inspires a model for most of the ordered portion of the toxic, full-length α-synuclein fibril, presenting opportunities for the design of inhibitors of α-synuclein fibrils.

Sigma-RF: prediction of the variability of spatial restraints in template-based modeling by random forest.

BackgroundIn template-based modeling when using a single template, inter-atomic distances of an unknown protein structure are assumed to be distributed by Gaussian probability density functions, whose center peaks are located at the distances between corresponding atoms in the template structure. The width of the Gaussian distribution, the variability of a spatial restraint, is closely related to the reliability of the restraint information extracted from a template, and it should be accurately estimated for successful template-based protein structure modeling.ResultsTo predict the variability of the spatial restraints in template-based modeling, we have devised a prediction model, Sigma-RF, by using the random forest (RF) algorithm. The benchmark results on 22 CASP9 targets show that the variability values from Sigma-RF are of higher correlations with the true distance deviation than those from Modeller. We assessed the effect of new sigma values by performing the single-domain homology modeling of 22 CASP9 targets and 24 CASP10 targets. For most of the targets tested, we could obtain more accurate 3D models from the identical alignments by using the Sigma-RF results than by using Modeller ones.ConclusionsWe find that the average alignment quality of residues located between and at two aligned residues, quasi-local information, is the most contributing factor, by investigating the importance of input features used in the RF machine learning. This average alignment quality is shown to be more important than the previously identified quantity of a local information: the product of alignment qualities at two aligned residues.Electronic supplementary materialThe online version of this article (doi:10.1186/s12859-015-0526-z) contains supplementary material, which is available to authorized users.