Protein Primary Sequence Research Articles

Reconnoitering the sequence and structural analysis of Staphylococcus aureus “A” protein

BackgroundThe Staphylococcus aureus “A” protein plays an essential role in the pathogenicity and virulence of this bacterial species. To gain deeper insights into the protein's characteristics, we conducted an in-depth analysis of its sequence and structure. ObjectiveThis study aimed to unravel the underlying genetic and structural components that contribute to the protein's functional properties. ResultsUtilizing various bioinformatics tools and techniques, we first examined the protein's primary sequence, identifying key amino acid residues and potential functional domains. Additionally, we employed computational modeling and simulation approaches to determine the tertiary structure of the “A” protein. Through this comprehensive analysis, we discovered novel features and interactions within the protein's structure, shedding light on its potential mechanisms of action. Furthermore, we investigated the protein's evolutionary conservation and compared it with related proteins from other bacterial species. ConclusionsOverall, our findings provide valuable insights into the sequence and structure of the Staphylococcus aureus “A” protein, which may have implications for understanding its role in pathogenicity and guiding the development of novel therapeutic strategies.

The M2 proteins of bat influenza A viruses reveal atypical features compared to conventional M2 proteins.

The influenza A virus (IAV) M2 protein has proton channel activity, which plays a role in virus uncoating and may help to preserve the metastable conformation of the IAV hemagglutinin (HA). In contrast to the highly conserved M2 proteins of conventional IAV, the primary sequences of bat IAV H17N10 and H18N11 M2 proteins show remarkable divergence, suggesting that these proteins may differ in their biological function. We, therefore, assessed the proton channel activity of bat IAV M2 proteins and investigated its role in virus replication. Here, we show that the M2 proteins of bat IAV did not fully protect acid-sensitive HA of classical IAV from low pH-induced conformational change, indicating low proton channel activity. Interestingly, the N31S substitution not only rendered bat IAV M2 proteins sensitive to inhibition by amantadine but also preserved the metastable conformation of acid-sensitive HA to a greater extent. In contrast, the acid-stable HA of H18N11 did not rely on such support by M2 protein. When mutant M2(N31S) protein was expressed in the context of chimeric H18N11/H5N1(6:2) encoding HA and NA of avian IAV H5N1, amantadine significantly inhibited virus entry, suggesting that ion channel activity supported virus uncoating. Finally, the cytoplasmic domain of the H18N11 M2 protein mediated rapid internalization of the protein from the plasma membrane leading to low-level expression at the cell surface. However, cell surface levels of H18N11 M2 protein were significantly enhanced in cells infected with the chimeric H18N11/H5N1(6:2) virus. The potential role of the N1 sialidase in arresting M2 internalization is discussed. IMPORTANCE Bat IAV M2 proteins not only differ from the homologous proteins of classical IAV by their divergent primary sequence but are also unable to preserve the metastable conformation of acid-sensitive HA, indicating low proton channel activity. This unusual feature may help to avoid M2-mediated cytotoxic effects and inflammation in bats infected with H17N10 or H18N11. Unlike classical M2 proteins, bat IAV M2 proteins with the N31S substitution mediated increased protection of HA from acid-induced conformational change. This remarkable gain of function may help to understand how single point mutations can modulate proton channel activity. In addition, the cytoplasmic domain was found to be responsible for the low cell surface expression level of bat IAV M2 proteins. Given that the M2 cytoplasmic domain of conventional IAV is well known to participate in virus assembly at the plasma membrane, this atypical feature might have consequences for bat IAV budding and egress.

Radical-Mediated Covalent Azidylation of Hydrophobic Microdomains in Water-Soluble Proteins.

Hydrophobic microdomains, also known as hydrophobic patches, are essential for many important biological functions of water-soluble proteins. These include ligand or substrate binding, protein-protein interactions, proper folding after translation, and aggregation during denaturation. Unlike transmembrane domains, which are easily recognized from stretches of contiguous hydrophobic sidechains in amino acids via primary protein sequence, these three-dimensional hydrophobic patches cannot be easily predicted. The lack of experimental strategies for directly determining their locations hinders further understanding of their structure and function. Here, we posit that the small triatomic anion N3- (azide) is attracted to these patches and, in the presence of an oxidant, forms a radical that covalently modifies C-H bonds of nearby amino acids. Using two model proteins (BSA and lysozyme) and a cell-free lysate from the model higher plant Arabidopsis thaliana, we find that radical-mediated covalent azidylation occurs within buried catalytic active sites and ligand binding sites and exhibits similar behavior to established hydrophobic probes. The results herein suggest a model in which the azido radical is acting as an "affinity reagent" for nonaqueous three-dimensional protein microenvironments and is consistent with both the nonlocalized electron density of the azide moiety and the known high reactivity of azido radicals widely used in organic chemistry syntheses. We propose that the azido radical is a facile means of identifying hydrophobic microenvironments in soluble proteins and, in addition, provides a simple new method for attaching chemical handles to proteins without the need for genetic manipulation or specialized reagents.

The distinct initiation sites and processing activities of TTLL4 and TTLL7 in glutamylation of brain tubulin

Mammalian brain tubulins undergo a reversible posttranslational modification-polyglutamylation-which attaches a secondary polyglutamate chain to the primary sequence of proteins. Loss of its erasers can disrupt polyglutamylation homeostasis and cause neurodegeneration. Tubulin tyrosine ligase like 4 (TTLL4) and TTLL7 were known to modify tubulins, both with preference for the β-isoform, but differently contribute to neurodegeneration. However, differences in their biochemical properties and functions remain largely unknown. Here, using an antibody-based method, we characterized the properties of a purified recombinant TTLL4 and confirmed its sole role as an initiator, unlike TTLL7, which both initiates and elongates the side chains. Unexpectedly, TTLL4 produced stronger glutamylation immunosignals for α-isoform than β-isoform in brain tubulins. Contrarily, the recombinant TTLL7 raised comparable glutamylation immunoreactivity for two isoforms. Given the site selectivity of the glutamylation antibody, we analyzed modification sites of two enzymes. Tandem mass spectrometry analysis revealed their incompatible site selectivity on synthetic peptides mimicking carboxyl termini of α1- and β2-tubulins and a recombinant tubulin. Particularly, in the recombinant α1A-tubulin, a novel region was found glutamylated by TTLL4 and TTLL7, that again at distinct sites. These results pinpoint different site specificities between two enzymes. Moreover, TTLL7 exhibits less efficiency to elongate microtubules premodified by TTLL4, suggesting possible regulation of TTLL7 elongation activity by TTLL4-initiated sites. Finally, we showed that kinesin behaves differentially on microtubules modified by two enzymes. This study underpins the different reactivity, site selectivity, and function of TTLL4 and TTLL7 on brain tubulins and sheds light on their distinct rolein vivo.

Exploiting conformational dynamics to modulate the function of designed proteins

With the recent success in calculating protein structures from amino acid sequences using artificial intelligence-based algorithms, an important next step is to decipher how dynamics is encoded by the primary protein sequence so as to better predict function. Such dynamics information is critical for protein design, where strategies could then focus not only on sequences that fold into particular structures that perform a given task, but would also include low-lying excited protein states that could influence the function of the designed protein. Herein, we illustrate the importance of dynamics in modulating the function of C34, a designed α/β protein that captures β-strands of target ligands and is a member of a family of proteins designed to sequester β-strands and β hairpins of aggregation-prone molecules that lead to a variety of pathologies. Using a strategy to "see" regions of apo C34 that are invisible to NMR spectroscopy as a result of pervasive conformational exchange, as well as a mutagenesis approach whereby C34 molecules are stabilized into a single conformer, we determine the structures of the predominant conformations that are sampled by C34 and show that these attenuate the affinity for cognate peptide. Subsequently, the observed motion is exploited to develop an allosterically regulated peptide binder whose binding affinity can be controlled through the addition of a second molecule. Our study emphasizes the unique role that NMR can play in directing the design process and in the construction of new molecules with more complex functionality.

Probing unfolding barriers in high throughput.

Energy barriers between protein states determine the rate at which a folded protein makes excursions to and from any high energy states. These may include functional states or potentially harmful states. Although barriers between states are important features of a protein's energy landscape, we do not understand how they are encoded in a protein's primary sequence. Therefore, we have developed an assay coupling quantitative proteolysis with a high throughput display system to probe the effect of sequence changes on a protein's unfolding barrier. Under certain conditions, proteolytic resistance arises solely from a large kinetic barrier to unfolding, allowing us to measure barrier heights directly. In addition to expanding our basic knowledge of protein folding, exploration of the relationship between sequence and kinetic barrier heights may contribute to a greater understanding of how pathogenic protein states arise and may aid to the development of long-lived, shelf-stable protein drugs.

CSM-Toxin: A Web-Server for Predicting Protein Toxicity

Biologics are one of the most rapidly expanding classes of therapeutics, but can be associated with a range of toxic properties. In small-molecule drug development, early identification of potential toxicity led to a significant reduction in clinical trial failures, however we currently lack robust qualitative rules or predictive tools for peptide- and protein-based biologics. To address this, we have manually curated the largest set of high-quality experimental data on peptide and protein toxicities, and developed CSM-Toxin, a novel in-silico protein toxicity classifier, which relies solely on the protein primary sequence. Our approach encodes the protein sequence information using a deep learning natural languages model to understand "biological" language, where residues are treated as words and protein sequences as sentences. The CSM-Toxin was able to accurately identify peptides and proteins with potential toxicity, achieving an MCC of up to 0.66 across both cross-validation and multiple non-redundant blind tests, outperforming other methods and highlighting the robust and generalisable performance of our model. We strongly believe the CSM-Toxin will serve as a valuable platform to minimise potential toxicity in the biologic development pipeline. Our method is freely available as an easy-to-use webserver.

Protein-Based Biological Materials: Molecular Design and Artificial Production.

Polymeric materials produced from fossil fuels have been intimately linked to the development of industrial activities in the 20th century and, consequently, to the transformation of our way of living. While this has brought many benefits, the fabrication and disposal of these materials is bringing enormous sustainable challenges. Thus, materials that are produced in a more sustainable fashion and whose degradation products are harmless to the environment are urgently needed. Natural biopolymers─which can compete with and sometimes surpass the performance of synthetic polymers─provide a great source of inspiration. They are made of natural chemicals, under benign environmental conditions, and their degradation products are harmless. Before these materials can be synthetically replicated, it is essential to elucidate their chemical design and biofabrication. For protein-based materials, this means obtaining the complete sequences of the proteinaceous building blocks, a task that historically took decades of research. Thus, we start this review with a historical perspective on early efforts to obtain the primary sequences of load-bearing proteins, followed by the latest developments in sequencing and proteomic technologies that have greatly accelerated sequencing of extracellular proteins. Next, four main classes of protein materials are presented, namely fibrous materials, bioelastomers exhibiting high reversible deformability, hard bulk materials, and biological adhesives. In each class, we focus on the design at the primary and secondary structure levels and discuss their interplays with the mechanical response. We finally discuss earlier and the latest research to artificially produce protein-based materials using biotechnology and synthetic biology, including current developments by start-up companies to scale-up the production of proteinaceous materials in an economically viable manner.

Design and Development of Hydrophobicity and Net charge Based Artificial Neural Network Model for IDP/IDPR Prediction

Intrinsically disordered proteins (IDPs) and proteins containing intrinsically disordered protein regions (IDPRs) often affect protein structure determination. IDP/IDPRs cause failures in the protein structure determination pipeline leading to enhanced experimental cost and time. Hence the primary sequence of proteins is often analyzed using disorder prediction servers so that manipulations can be made to the protein sequence to aid its expression, purification, and crystallization. Assessment of existing IDP/IDPR prediction methods with CASP 10 targets shows the scope of improvement of prediction. In this paper, an attempt has been made to develop a hydrophobicity and net charge based artificial neural network model for protein disorder prediction. Subsequently, a prediction tool has been developed using Java using the Neural Network Schema. The prediction tool has been tested with CASP 10 targets and the developed predictor is found to outperform other predictors i.e RONN, PONDR VLXT, DisEMBL, FOLDINDEX, and GLOBPLOT in Sensitivity (0.983), Precision (0.854), Accuracy (0.937), MCC (0.840), and AUC (0.937). DisEMBL and GLOBPLOT are found to have better specificity i.e 0.909 and 0.979 respectively than the proposed predictor i.e 0.890.

EnsembleDL-ATG: Identifying autophagy proteins by integrating their sequence and evolutionary information using an ensemble deep learning framework

Autophagy is a primary mechanism for maintaining cellular homeostasis. The synergistic actions of autophagy-related (ATG) proteins strictly regulate the whole autophagic process. Therefore, accurate identification of ATGs is a first and critical step to reveal the molecular mechanism underlying the regulation of autophagy. Current computational methods can predict ATGs from primary protein sequences, but owing to the limitations of algorithms, significant room for improvement still exists. In this research, we propose EnsembleDL-ATG, an ensemble deep learning framework that aggregates multiple deep learning models to predict ATGs from protein sequence and evolutionary information. We first evaluated the performance of individual networks for various feature descriptors to identify the most promising models. Then, we explored all possible combinations of independent models to select the most effective ensemble architecture. The final framework was built and maintained by an organization of four different deep learning models. Experimental results show that our proposed method achieves a prediction accuracy of 94.5 % and MCC of 0.890, which are nearly 4 % and 0.08 higher than ATGPred-FL, respectively. Overall, EnsembleDL-ATG is the first ATG machine learning predictor based on ensemble deep learning. The benchmark data and code utilized in this study can be accessed for free at https://github.com/jingry/autoBioSeqpy/tree/2.0/examples/EnsembleDL-ATG.

GUD-VE visualization tool for physicochemical properties of proteins

The physicochemical properties of primary sequences of proteins helps in determining both the structure and biological functions. The sequence analysis of the proteins and nucleic acids is most fundamental element of bioinformatics. Without these elements, it is impossible to gain insight deeper molecular and biochemical mechanisms. For this purpose, the computational methods like bioinformatics tools assist experts and novices alike in resolving issues relating to protein analysis. Similarly, this proposed work, for the graphical user interface (GUI) based prediction and visualization through the computations-based method done on Jupyter Notebook with tkinter package which allows the creation of a program on a local host platform and accessed by the programmer.•When it is queried with a protein sequence, it predicts physicochemical parameters of the peptides.•Users can choose to visualize the findings acquired either anonymously or on the user-specified email address and compare the biophysical properties of one protein with other using amino acids (AA) sequences.The aim of this paper is to meet the requirements of experimentalists, not just hardcore bioinformaticians related to biophysical properties prediction and comparison with other proteins. The code for it has been uploaded on GitHub (an online repository of codes) in private mode.

Online Cross-Linking of Peptides and Proteins during Contained-Electrospray Ionization Mass Spectrometry.

Recent advancements in mass spectrometry (MS) now enable all levels of protein structures to be characterized, including primary protein sequence, post-translational modifications, and three-dimensional protein conformations. However, protein conformational studies by MS require the use of many separate techniques that are performed independently of each other. Herein, we described a contained-electrospray (ES) experiment that has potential to integrate peptide/protein cross-linking with the general MS workflow. In our experiment, cross-linking of protein/peptide occurs simultaneously with ionization after analytes, and cross-linkers are sprayed from two separate ES emitters. The online cross-linking process occurring in the charged microdroplet environment was optimized using trilysine peptide and bis(sulfosuccinimidyl)suberate cross-linker. We detected the electrostatic complex between analyte and cross-linker, the mono-linked intermediate, and the fully cross-linked product, allowing us to correctly predict the sequence of reaction events in the cross-linking process. Importantly, we observed that the terminal fully cross-linked product is composed of two distinct conformations. In one form, the product involved cross-linking between two ε-NH2 amines in lysine residues, while the other conformer was formed by a reaction between one ε-NH2 amine and the N-terminus. The experimental conditions for selecting one cross-linked species over others during the online ES ionization-MS analysis have been detailed. Appropriate parameters enabled the reaction between α-lactalbumin proteins and cross-linkers using a non-denaturing spray condition. These results establish a framework for a future development in high-throughput structural MS method, where all levels of protein information can be gathered in a single experiment.

Improved compound–protein interaction site and binding affinity prediction using self-supervised protein embeddings

BackgroundCompound–protein interaction site and binding affinity predictions are crucial for drug discovery and drug design. In recent years, many deep learning-based methods have been proposed for predications related to compound–protein interaction. For protein inputs, how to make use of protein primary sequence and tertiary structure information has impact on prediction results.ResultsIn this study, we propose a deep learning model based on a multi-objective neural network, which involves a multi-objective neural network for compound–protein interaction site and binding affinity prediction. We used several kinds of self-supervised protein embeddings to enrich our protein inputs and used convolutional neural networks to extract features from them. Our results demonstrate that our model had improvements in terms of interaction site prediction and affinity prediction compared to previous models. In a case study, our model could better predict binding sites, which also showed its effectiveness.ConclusionThese results suggest that our model could be a helpful tool for compound–protein related predictions.

Robust and accurate prediction of self-interacting proteins from protein sequence information by exploiting weighted sparse representation based classifier

BackgroundSelf-interacting proteins (SIPs), two or more copies of the protein that can interact with each other expressed by one gene, play a central role in the regulation of most living cells and cellular functions. Although numerous SIPs data can be provided by using high-throughput experimental techniques, there are still several shortcomings such as in time-consuming, costly, inefficient, and inherently high in false-positive rates, for the experimental identification of SIPs even nowadays. Therefore, it is more and more significant how to develop efficient and accurate automatic approaches as a supplement of experimental methods for assisting and accelerating the study of predicting SIPs from protein sequence information.ResultsIn this paper, we present a novel framework, termed GLCM-WSRC (gray level co-occurrence matrix-weighted sparse representation based classification), for predicting SIPs automatically based on protein evolutionary information from protein primary sequences. More specifically, we firstly convert the protein sequence into Position Specific Scoring Matrix (PSSM) containing protein sequence evolutionary information, exploiting the Position Specific Iterated BLAST (PSI-BLAST) tool. Secondly, using an efficient feature extraction approach, i.e., GLCM, we extract abstract salient and invariant feature vectors from the PSSM, and then perform a pre-processing operation, the adaptive synthetic (ADASYN) technique, to balance the SIPs dataset to generate new feature vectors for classification. Finally, we employ an efficient and reliable WSRC model to identify SIPs according to the known information of self-interacting and non-interacting proteins.ConclusionsExtensive experimental results show that the proposed approach exhibits high prediction performance with 98.10% accuracy on the yeast dataset, and 91.51% accuracy on the human dataset, which further reveals that the proposed model could be a useful tool for large-scale self-interacting protein prediction and other bioinformatics tasks detection in the future.

Predicting emerging SARS-CoV-2 variants of concern through a One Class dynamic anomaly detection algorithm

ObjectivesThe objective of this study is the implementation of an automatic procedure to weekly detect new SARS-CoV-2 variants and non-neutral variants (variants of concern (VOC) and variants of interest (VOI)).MethodsWe...

Silencing of HN1L suppresses the proliferation and migration of cancer cells

Background and purpose: HN1L is a member of the HN1 gene family and shares about 30% similarity with HN1 which is another member of the family on the primary protein sequence. Since HN1 is an important gene that is involved in various cellular mechanisms and also differentially expressed in carcinogenesis, we investigated the effect of HN1L on some malignant behaviors of various cancer cells.Material and methods: Co-expression analysis, Gene Ontology enrichment, and database searches were performed to predict the cellular roles of HN1, and to investigate its expression in cancers and their corresponding normal tissues. Western blotting and Real-Time PCR were used to compare the expression of HN1L in the normal prostate cells and prostate cancer cells. Cell proliferation and migration assays were used to investigate the effects of HN1L depletion on cell proliferation and migration.Results: The results of co-expression and Gene Ontology enrichment analyses showed that HN1L is co-expressed with DNA replication and DNA damage response/repair associated genes. The database search results revealed that HN1L expression increases in at least 10 diverse cancer types compared to their normal corresponding tissues. This result was confirmed in the prostate cancer cell model, experimentally. Silencing of HN1L inhibited proliferative and migrative behaviors of prostate, breast, colon, and cervix cancer cells.Conclusions: HN1L probably is a novel proto-oncogene that is involved in the DNA metabolism-related mechanisms, and high HN1L level promotes further proliferation and migration in the cancer cells.

Protein encoder: An autoencoder-based ensemble feature selection scheme to predict protein secondary structure

Proteins play a vital role in the human body as they perform important metabolic tasks. Experimental identification of protein structure is expensive and time consuming. The prediction of protein secondary structure is significant to identify the protein tertiary structure and its folds. The feature subset selection from high dimensional protein primary sequence is a key to improve the accuracy of Protein Secondary Structure Prediction (PSSP). Therefore, it is essential to select the relevant features from high dimensional data to predict the protein secondary structure. This work presents a novel method for the PSSP problem based on a two-phase feature selection technique. The first stage utilizes an unsupervised autoencoder for feature extractions. Whereas, the second stage is an ensemble of three feature selection methods, namely, generic univariate select, recursive feature elimination, and Pearson's correlation. This phase combines multiple feature subsets using mutual information to select the optimum feature subset. For classification, different resultant subset features are used. These include random forest, decision tree, and multilayer perceptron. Two sets of experiments are performed on five datasets for the assessment of proposed work. The proposed solution is compared with three state-of-the-art methods based on Q3 accuracy, Q8 accuracy, and segment overlap score. Obtained results show that the proposed framework performs better in the majority of the cases than the past contributions. The proposed framework achieves Q8 accuracies of 82%, 80%, 79%, 73%, and 74% and Q3 accuracies of 90%, 90%, 92%, 79%, and 74% on CB6133, CB6133-filtered, CB513, CASP10, and CASP11 datasets, respectively.

Differential glycosylation of tissue non-specific alkaline phosphatase in mesenchymal stromal cells differentiated into either an osteoblastic or adipocytic phenotype

It has long been known that tissue non-specific alkaline phosphatase (TNAP) is essential for the correct formation of bone, as altered expression or function of this enzyme results in hypophosphatasia, a disease characterised by compromised bone structure, density and strength. However, recent evidence strongly suggests that the enzyme also has a role in lipid accrual and adipogenesis, a function that seems far removed from bone formation. Given that mesenchymal stromal cells (MSCs) are progenitors of both osteoblasts and adipocytes, the question arises of how TNAP is regulated to potentially have a different function when MSCs undergo either osteogenesis or adipogenesis. As the primary protein sequence is unchanged for the enzyme during both types of differentiation, any differences in function must be attributed to post-translational modification and/or localisation. We therefore examined the location of TNAP in bone- or adipose-derived MSCs differentiated into an adipocytic phenotype and compared the glycosylation state of the enzyme in MSCs differentiated into either osteoblasts or adipocytes. TNAP was found to co-locate with perilipin around lipid droplets in MSCs from bone, subcutaneous- and visceral adipose tissue during adipocytic differentiation. Treatment of TNAP with wheat germ lectin followed by electrophoresis showed minor differences in glycosylation between the phosphatase isolated from cells from these tissues, whereas electrophoresis after neuraminidase digestion highlighted differential glycosylation between cell types and during adipogenesis and osteoblastogenesis. This infers that post-translational modification of TNAP is altered during differentiation and is dependent on the eventual phenotype of the cells.

In Silico Epitope-Based Vaccine Prediction against Fungal Infection Aspergillosis

Aspergillus fumigatus is a pathogenic microorganism that causes aspergillosis due to the presence of its allergenic proteins. During the last two years, a few clinical cases have been reported where allergic bronchopulmonary aspergillosis (ABPA) has been detected in COVID-19 patients. The administration of antifungal medicine did not provide satisfactory results. It is a challenging job for medical scientists to protect mankind by designing an epitope-based vaccine against the rare disease aspergillosis. Other than twenty-three allergenic proteins, this microorganism contains an extra-cellular cellulase CelA expansin protein (Afu5g08030), which is allergenic. To design a peptide vaccine against aspergillosis, the identification of B cell and T cell epitopes is state-of-the-art technology. In our latest research, probable T cell and B cell epitopes are predicted. Molecular docking analysis of these predicted epitopes with their receptors is performed. Here, the primary sequence of the expansin protein is extracted and analyzed. Then, its secondary and tertiary structures are predicted using a homology modeling method and validated. Considering the physicochemical properties of this antigenic protein, two short stretches of peptides, namely 80KPQADEDPNASSSSSSS96 and 286DGGKTWQGTTRTS298, are predicted as linear B cell epitopes. Similarly, based on its contacts with the highest number of alleles, the peptide sequence 221LDLFQNAFTQLADVS235 is chosen as the most possible T cell epitope for the protein present in Aspergillus fumigatus with the highest binding energy for MHC II allele HLA-DRB1* 01: 01. Considering the binding energy of the B cell epitope with IgE, the second epitope 286DGGKTWQGTTRTS298 is designated as the most potential epitope of B cells for this protein. Docking studies were performed with the T cell epitope with the human ternary complex of T cell receptor, CD4 receptor, and peptide-MHC II molecule (PDB ID 3T0E) with a binding energy of −192 Kcal/mole. For peptide-based vaccines, the proposed B cell and T cell epitopes may be used against aspergillosis after further experimental analysis.

Interpretable machine learning identification of arginine methylation sites

Protein methylation is one of the most prominent posttranslation modifications that essentially regulates several biological processes in eukaryotes. Therefore, identification of the arginine methylation site is crucial in deciphering its characteristics and functions in cell biology, disease mechanisms, and guided drug development. The computation methods address the long-term bottleneck together with the cost, time, and labor required in experimental methods for large-scale identification of protein arginine methylation sites. In this study, we proposed a robust machine learning-based computational tool known as iIRMethyl, employing the primary sequence and physicochemical properties of protein along with a two-step feature selection method for optimal selection of feature descriptors. Moreover, the performance of iIRMethyl was comprehensively evaluated via k-fold cross-validation on a benchmark dataset and independent test dataset. iIRMethyl demonstrated a remarkably greater performance than the state-of-the-art method and achieved an average area under the curve value of 0.99 for both k-fold cross-validation and an independent test set in the identification of protein arginine methylation sites. Furthermore, the outcomes reveal that iIRMethyl is a robust and accurate computational tool for large-scale identification of arginine methylation sites and would facilitate the understanding of their functional mechanisms and accelerating their application in drug development and clinical therapy. Additionally, the prediction mechanism of the proposed model iIRMethyl is interpreted using the SHapley Additive exPlanation algorithm.