Common Structural Pattern Research Articles

Analysis of the abundance and diversity of RNA secondary structure elements in RNA viruses using the RNAsselem Python package

Recent advancements in experimental and computational methods for RNA secondary structure detection have revealed the crucial role of RNA structural elements in diverse molecular processes within living cells. It has been demonstrated that the secondary structure of the entire viral genome is often responsible for performing crucial functions in the viral life cycle and also influences virus evolution. To investigate the role of viral RNA secondary structure, alongside experimental techniques, the use of bioinformatics tools is important for analyzing various secondary structure patterns, including hairpin loops, internal loops, multifurcations, external loops, bulges, stems, and pseudoknots. Here, we have introduced a Python package for analyzing RNA secondary structure elements in viral genomes, which includes the recognition of common secondary structure patterns, the generation of descriptive statistics for these structural elements, and the provision of their basic properties. We applied the developed package to analyze the secondary structures of complete viral genomes collected from the literature, aiming to gain insights into viral function and evolution. Both the package and the collection of secondary structures of viral genomes are available at http://github.com/KazanovLab/RNAsselem.

From Double-Stranded Helicates to Abiotic Double Helical Supramolecular Assemblies.

The folding of oligomeric strands is the method that nature has selected to generate ordered assemblies presenting spectacular functions. In the purpose to mimic these biomacro-molecules and extend their properties and functions, chemists make important efforts to prepare artificial secondary, tertiary, and even quarternary structures based on folded abiotic backbones. A large variety of oligomers and polymers, encoded with chemical informations, were designed, synthesized and characterized, and the establishment of non-covalent interactions lead to complex and functional supramolecular architectures resulting from a spontaneous self-assembly process. The association of comple-mentary molecular strands into double helical structures is a common structural pattern of nucleic acids and proteins, so the synthesis of bio-inspired double helices has emerged as an important subject. In recent years, a number of synthetic oligo-mers have been reported to form stable double helices and it was shown that the equilibrium between single and double helices can be controlled via different stimuli like the modification of the solvent or the temperature. This kind of structure presents highly interesting functions, such as molecular recognition within the cavity of double helices, and some other potential applications will emerge in the future.

DOMICO: Checking conformance between domain models and implementations

AbstractAs a predominant design method for microsservices architecture (MSA), domain‐driven design (DDD) utilizes a series of standard patterns in both models and implementations to effectively support the design of architectural elements. However, an implementation may deviate from its original domain model that uses certain patterns. The deviation between a domain model and its implementation is a type of architectural drift, which needs to be detected promptly. This paper proposes an approach, namely DOMICO, to check the conformance between the domain model and its implementation, by which the conformance is formalized by defining eight common structural patterns of domain modeling and their representations in both models and the corresponding source code. Based on the formalization, our approach can not only identify the discrepancies (e.g., divergence, absence, and modification) with respect to pattern elements, but also detect possible violations of 24 compliance rules imposed by the patterns. To validate DOMICO, we performed a case study to investigate its use in a supply chain project and its performance. The results show that DOMICO can accurately identify 100% inconsistency issues in the cases examined. As the first conformance checking approach for DDD, DOMICO can be integrated into the regular domain modeling process and help ensure the conformity of microservice implementations to models.

Doxorubicin Changes the Spatial Organization of the Genome around Active Promoters.

In this study, we delve into the impact of genotoxic anticancer drug treatment on the chromatin structure of human cells, with a particular focus on the effects of doxorubicin. Using Hi-C, ChIP-seq, and RNA-seq, we explore the changes in chromatin architecture brought about by doxorubicin and ICRF193. Our results indicate that physiologically relevant doses of doxorubicin lead to a local reduction in Hi-C interactions in certain genomic regions that contain active promoters, with changes in chromatin architecture occurring independently of Top2 inhibition, cell cycle arrest, and differential gene expression. Inside the regions with decreased interactions, we detected redistribution of RAD21 around the peaks of H3K27 acetylation. Our study also revealed a common structural pattern in the regions with altered architecture, characterized by two large domains separated from each other. Additionally, doxorubicin was found to increase CTCF binding in H3K27 acetylated regions. Furthermore, we discovered that Top2-dependent chemotherapy causes changes in the distance decay of Hi-C contacts, which are driven by direct and indirect inhibitors. Our proposed model suggests that doxorubicin-induced DSBs cause cohesin redistribution, which leads to increased insulation on actively transcribed TAD boundaries. Our findings underscore the significant impact of genotoxic anticancer treatment on the chromatin structure of the human genome.

Prediction of chaperonin GroE substrates using small structural patterns of proteins.

Molecular chaperones are indispensable proteins that assist the folding of aggregation-prone proteins into their functional native states, thereby maintaining organized cellular systems. Two of the best characterized chaperones are the Escherichia coli chaperonins GroEL and GroES (GroE), for which in vivo obligate substrates have been identified by proteome-wide experiments. These substrates comprise various proteins, but exhibit remarkable structural features. They include a number of α/β proteins, particularly those adopting the TIM β/α barrel fold. This observation led us to speculate that GroE obligate substrates share a structural motif. Based on this hypothesis, we exhaustively compared substrate structures with the MICAN alignment tool, which detects common structural patterns while ignoring the connectivity or orientation of secondary structural elements. We selected four (or five) substructures with hydrophobic indices that were mostly included in substrates and excluded in others, and developed a GroE obligate substrate discriminator. The substructures are structurally similar and superimposable on the 2-layer 2α4β sandwich, the most popular protein substructure, implying that targeting this structural pattern is a useful strategy for GroE to assist numerous proteins. Seventeen false positives predicted by our methods were experimentally examined using GroE-depleted cells, and 9 proteins were confirmed to be novel GroE obligate substrates. Together, these results demonstrate the utility of our common-substructure hypothesis and prediction method.

Maritime Students Meeting the Maritime Industry English Standards: An Analysis of Types of Sentences

A sentence is the highest unit of grammar. Thus, constructing error-free sentences in writing is one of the biggest challenges encounted by most non-native speakers, and even university students are not an exception to this reality. This study aims at investigating various types of sentences produced by tertiary-level maritime students in a Sri Lankan university. The study was based on a narrative writing activity in the English for Academic Purposes (EAP) module. The students were provided forty-five minutes to produce the piece of writing as an in-class activity on a topic relevant to their field visit to a port. This is a descriptive study based on the analysis of a small corpus of essays written by twenty maritime students, and a structural analysis of sentences was employed to examine the students’ writing. Different kinds of sentences and sentence errors were identified, and they were classified accordingly. The findings of the study revealed that the students favoured simple sentences over other sentence types. Approximately two-thirds of the sentences produced belonged to the simple sentence category. The compound-complex form was found to be the least utilized sentence type among the target group. The analysis of sentences was based on the elements of the clause structure explained in Quirk et al. (1985) and Oshima and Hough (2006). Interestingly, it was observed that there was no single common clause structural pattern used by the participants. Instead, they used subject-verb-object (SVO), subject-verb-complement (SVC) and subject-verb-adverbial (SVA) types very often when writing. Similarly, fragments and run-on sentences were recorded high among maritime learners’ erroneous sentences in writing. The study findings have pedagogical implications for the teaching of English language grammar that subsumes essay writing in the EAP module. 

Analysis of Host–Bacteria Protein Interactions Reveals Conserved Domains and Motifs That Mediate Fundamental Infection Pathways

Adhesion and colonization of host cells by pathogenic bacteria depend on protein–protein interactions (PPIs). These interactions are interesting from the pharmacological point of view since new molecules that inhibit host-pathogen PPIs would act as new antimicrobials. Most of these interactions are discovered using high-throughput methods that may display a high false positive rate. The absence of curation of these databases can make the available data unreliable. To address this issue, a comprehensive filtering process was developed to obtain a reliable list of domains and motifs that participate in PPIs between bacteria and human cells. From a structural point of view, our analysis revealed that human proteins involved in the interactions are rich in alpha helix and disordered regions and poorer in beta structure. Disordered regions in human proteins harbor short sequence motifs that are specifically recognized by certain domains in pathogenic proteins. The most relevant domain–domain interactions were validated by AlphaFold, showing that a proper analysis of host-pathogen PPI databases can reveal structural conserved patterns. Domain–motif interactions, on the contrary, were more difficult to validate, since unstructured regions were involved, where AlphaFold could not make a good prediction. Moreover, these interactions are also likely accommodated by post-translational modifications, especially phosphorylation, which can potentially occur in 25–50% of host proteins. Hence, while common structural patterns are involved in host–pathogen PPIs and can be retrieved from available databases, more information is required to properly infer the full interactome. By resolving these issues, and in combination with new prediction tools like Alphafold, new classes of antimicrobials could be discovered from a more detailed understanding of these interactions.

Deep learning methods may not outperform other machine learning methods on analyzing genomic studies.

Deep Learning (DL) has been broadly applied to solve big data problems in biomedical fields, which is most successful in image processing. Recently, many DL methods have been applied to analyze genomic studies. However, genomic data usually has too small a sample size to fit a complex network. They do not have common structural patterns like images to utilize pre-trained networks or take advantage of convolution layers. The concern of overusing DL methods motivates us to evaluate DL methods’ performance versus popular non-deep Machine Learning (ML) methods for analyzing genomic data with a wide range of sample sizes. In this paper, we conduct a benchmark study using the UK Biobank data and its many random subsets with different sample sizes. The original UK Biobank data has about 500k participants. Each patient has comprehensive patient characteristics, disease histories, and genomic information, i.e., the genotypes of millions of Single-Nucleotide Polymorphism (SNPs). We are interested in predicting the risk of three lung diseases: asthma, COPD, and lung cancer. There are 205,238 participants have recorded disease outcomes for these three diseases. Five prediction models are investigated in this benchmark study, including three non-deep machine learning methods (Elastic Net, XGBoost, and SVM) and two deep learning methods (DNN and LSTM). Besides the most popular performance metrics, such as the F1-score, we promote the hit curve, a visual tool to describe the performance of predicting rare events. We discovered that DL methods frequently fail to outperform non-deep ML in analyzing genomic data, even in large datasets with over 200k samples. The experiment results suggest not overusing DL methods in genomic studies, even with biobank-level sample sizes. The performance differences between DL and non-deep ML decrease as the sample size of data increases. This suggests when the sample size of data is significant, further increasing sample sizes leads to more performance gain in DL methods. Hence, DL methods could be better if we analyze genomic data bigger than this study.

Cancer glycomics offers potential biomarkers and therapeutic targets in the framework of 3P medicine.

Glycosylation is one of the most important post-translational modifications (PTMs) in a protein, and is the most abundant and diverse biopolymer in nature. Glycans are involved in multiple biological processes of cancer initiation and progression, including cell-cell interactions, cell-extracellular matrix interactions, tumor invasion and metastasis, tumor angiogenesis, and immune regulation. As an important biomarker, tumor-associated glycosylation changes have been extensively studied. This article reviews recent advances in glycosylation-based biomarker research, which is useful for cancer diagnosis and prognostic assessment. Truncated O-glycans, sialylation, fucosylation, and complex branched structures have been found to be the most common structural patterns in malignant tumors. In recent years, immunochemical methods, lectin recognition-based methods, mass spectrometry (MS)-related methods, and fluorescence imaging-based in situ methods have greatly promoted the discovery and application potentials of glycomic and glycoprotein biomarkers in various cancers. In particular, MS-based proteomics has significantly facilitated the comprehensive research of extracellular glycoproteins, increasing our understanding of their critical roles in regulating cellular activities. Predictive, preventive and personalized medicine (PPPM; 3P medicine) is an effective approach of early prediction, prevention and personalized treatment for different patients, and it is known as the new direction of medical development in the 21st century and represents the ultimate goal and highest stage of medical development. Glycosylation has been revealed to have new diagnostic, prognostic, and even therapeutic potentials. The purpose of glycosylation analysis and utilization of biology is to make a fundamental change in health care and medical practice, so as to lead medical research and practice into a new era of 3P medicine.

Common Structural Pattern for Flecainide Binding in Atrial-Selective Kv1.5 and Nav1.5 Channels: A Computational Approach.

Atrial fibrillation (AF) is the most common cardiac arrhythmia. Its treatment includes antiarrhythmic drugs (AADs) to modulate the function of cardiac ion channels. However, AADs have been limited by proarrhythmic effects, non-cardiovascular toxicities as well as often modest antiarrhythmic efficacy. Theoretical models showed that a combined blockade of Nav1.5 (and its current, INa) and Kv1.5 (and its current, IKur) ion channels yield a synergistic anti-arrhythmic effect without alterations in ventricles. We focused on Kv1.5 and Nav1.5 to search for structural similarities in their binding site (BS) for flecainide (a common blocker and widely prescribed AAD) as a first step for prospective rational multi-target directed ligand (MTDL) design strategies. We present a computational workflow for a flecainide BS comparison in a flecainide-Kv1.5 docking model and a solved structure of the flecainide-Nav1.5 complex. The workflow includes docking, molecular dynamics, BS characterization and pattern matching. We identified a common structural pattern in flecainide BS for these channels. The latter belongs to the central cavity and consists of a hydrophobic patch and a polar region, involving residues from the S6 helix and P-loop. Since the rational MTDL design for AF is still incipient, our findings could advance multi-target atrial-selective strategies for AF treatment.

Paper writing: Basic requirements and common structural patterns of title

Paper writing: Basic requirements and common structural patterns of title

Mining of structural motifs in proteins using artificial bee colony optimization framework for druggability.

In this work, we have developed an optimization framework for digging out common structural patterns inherent in DNA binding proteins. A novel variant of the artificial bee colony optimization algorithm is proposed to improve the exploitation process. Experiments on four benchmark objective functions for different dimensions proved the speedier convergence of the algorithm. Also, it has generated optimum features of Helix Turn Helix structural pattern based on the objective function defined with occurrence count on secondary structure. The proposed algorithm outperformed the compared methods in convergence speed and the quality of generated motif features. The motif locations obtained using the derived common pattern are compared with the results of two other motif detection tools. 92% of tested proteins have produced matching locations with the results of the compared methods. The performance of the approach was analyzed with various measures and observed higher sensitivity, specificity and area under the curve values. A novel strategy for druggability finding by docking studies, targeting the motif locations is also discussed.

Identification and evaluation of potential SARS-CoV-2 antiviral agents targeting mRNA cap guanine N7-Methyltransferase

SARS-CoV-2, the cause of the currently ongoing COVID-19 pandemic, encodes its own mRNA capping machinery. Insights into this capping system may provide new ideas for therapeutic interventions and drug discovery. In this work, we employ a previously developed Py-FLINT screening approach to study the inhibitory effects of compounds against the cap guanine N7-methyltransferase enzyme, which is involved in SARS-CoV-2 mRNA capping. We screened five commercially available libraries (7039 compounds in total) to identify 83 inhibitors with IC50 < 50 μM, which were further validated using RP HPLC and dot blot assays. Novel fluorescence anisotropy binding assays were developed to examine the targeted binding site. The inhibitor structures were analyzed for structure-activity relationships in order to define common structural patterns. Finally, the most potent inhibitors were tested for antiviral activity on SARS-CoV-2 in a cell based assay.

ShapeMOD

A popular way to create detailed yet easily controllable 3D shapes is via procedural modeling, i.e. generating geometry using programs. Such programs consist of a series of instructions along with their associated parameter values. To fully realize the benefits of this representation, a shape program should be compact and only expose degrees of freedom that allow for meaningful manipulation of output geometry. One way to achieve this goal is to design higher-level macro operators that, when executed, expand into a series of commands from the base shape modeling language. However, manually authoring such macros, much like shape programs themselves, is difficult and largely restricted to domain experts. In this paper, we present ShapeMOD, an algorithm for automatically discovering macros that are useful across large datasets of 3D shape programs. ShapeMOD operates on shape programs expressed in an imperative, statement-based language. It is designed to discover macros that make programs more compact by minimizing the number of function calls and free parameters required to represent an input shape collection. We run ShapeMOD on multiple collections of programs expressed in a domain-specific language for 3D shape structures. We show that it automatically discovers a concise set of macros that abstract out common structural and parametric patterns that generalize over large shape collections. We also demonstrate that the macros found by ShapeMOD improve performance on downstream tasks including shape generative modeling and inferring programs from point clouds. Finally, we conduct a user study that indicates that ShapeMOD's discovered macros make interactive shape editing more efficient.

Using the antibody-antigen binding interface to train image-based deep neural networks for antibody-epitope classification.

High-throughput B-cell sequencing has opened up new avenues for investigating complex mechanisms underlying our adaptive immune response. These technological advances drive data generation and the need to mine and analyze the information contained in these large datasets, in particular the identification of therapeutic antibodies (Abs) or those associated with disease exposure and protection. Here, we describe our efforts to use artificial intelligence (AI)-based image-analyses for prospective classification of Abs based solely on sequence information. We hypothesized that Abs recognizing the same part of an antigen share a limited set of features at the binding interface, and that the binding site regions of these Abs share share common structure and physicochemical property patterns that can serve as a "fingerprint" to recognize uncharacterized Abs. We combined large-scale sequence-based protein-structure predictions to generate ensembles of 3-D Ab models, reduced the Ab binding interface to a 2-D image (fingerprint), used pre-trained convolutional neural networks to extract features, and trained deep neural networks (DNNs) to classify Abs. We evaluated this approach using Ab sequences derived from human HIV and Ebola viral infections to differentiate between two Abs, Abs belonging to specific B-cell family lineages, and Abs with different epitope preferences. In addition, we explored a different type of DNN method to detect one class of Abs from a larger pool of Abs. Testing on Ab sets that had been kept aside during model training, we achieved average prediction accuracies ranging from 71-96% depending on the complexity of the classification task. The high level of accuracies reached during these classification tests suggests that the DNN models were able to learn a series of structural patterns shared by Abs belonging to the same class. The developed methodology provides a means to apply AI-based image recognition techniques to analyze high-throughput B-cell sequencing datasets (repertoires) for Ab classification.

ПРОГНОЗИРОВАНИЕ УЩЕРБОВ ОТ ПРИРОДНЫХ ОПАСНОСТЕЙ ДЛЯ ТУРИСТИЧЕСКОГО КЛАСТЕРА «КУРОРТЫ СЕВЕРНОГО КАВКАЗА»

The article proposes a methodological approach to potential damage from natural hazards forecasting in case of large-scale investment projects realization in ski tourism planning, as well as to assessing changes in the vulnerability of the territory in which these projects will be implemented. The method was verified on the data of the “Northern Caucasus Resorts” tourist cluster. The study purpose is the creation and verification of a methodology for socio-economic damage predicting in limit values and vulnerability changing in the regions of the “North Caucasus Resorts” tourist cluster objects localization for the long term. Research methods – statistical (a structural approach based on the identification of common structural patterns of several sets). The lack of statistical information on significant parameters for forecasting determine necessitates of using the various logically non-contradictory revaluations based on the identified structural similarities for the calculation of their values within the planning horizon. The study results and main conclusions – in case of the “North Caucasus Resorts” tourist cluster creation the number of people potentially located in avalanche and mudflow danger areas will significantly increase in all of its facilities localization municipalities, which indicate an increase in the individual risk of death level for this territory. The present population in the ski season in some of the most remote and underdeveloped areas can increase up to 30 times. The increment in the value of the fixed assets for the municipalities under consideration will be from two to 90 times, potential damages in limit values will reach tens of billions rubles.

GraLSP: Graph Neural Networks with Local Structural Patterns

It is not until recently that graph neural networks (GNNs) are adopted to perform graph representation learning, among which, those based on the aggregation of features within the neighborhood of a node achieved great success. However, despite such achievements, GNNs illustrate defects in identifying some common structural patterns which, unfortunately, play significant roles in various network phenomena. In this paper, we propose GraLSP, a GNN framework which explicitly incorporates local structural patterns into the neighborhood aggregation through random anonymous walks. Specifically, we capture local graph structures via random anonymous walks, powerful and flexible tools that represent structural patterns. The walks are then fed into the feature aggregation, where we design various mechanisms to address the impact of structural features, including adaptive receptive radius, attention and amplification. In addition, we design objectives that capture similarities between structures and are optimized jointly with node proximity objectives. With the adequate leverage of structural patterns, our model is able to outperform competitive counterparts in various prediction tasks in multiple datasets.

Structural and developmental patterns in the pericarp of Malpighiaceae: new data from winged fruits of Janusia, Mascagnia, and Tetrapterys

Malpighiaceae have highly varied fruits, and their characteristics are traditionally used in taxa delimitation. However, fruit characters are homoplastic; therefore, detailed analyses are indispensable. Here, we have described winged-fruit ontogeny and anatomy of four Malpighiaceae species from different clades, and compared them with others described in the literature, with the goal of recognizing common developmental and structural patterns. Flowers and fruits were collected and processed for light microscopy. In the ovary, the main patterns were a uniseriate outer epidermis, a mesophyll with two or three distinct regions, and an inner epidermis formed by elongated or isodiametric cells. In the pericarp, the uniseriate exocarp appeared to be consistent throughout the family, and the outermost parenchymatic layers of the mesocarp and the inner layers formed by thickened-lignified cells prevailed. Some species also had a middle layer in the mesocarp, composed of lignified cells. The most substantial variation was in the endocarp and was independent of fruit type and phylogenetic position. Three structural patterns were recognized: sclerenchymatous cells, elongated and non-lignified cells, and a seed cushion. These data allow for the identification of important structural patterns in Malpighiaceae, and the high structural variation in the endocarp is promising for its systematics.

Multi-View Low-Rank Coding-Based Network Data De-Anonymization

Social networks are extensively exploited by third-party consumers such as researchers and advertisers to understand user characteristics and behaviors. In general, before network data is published, sensitive relationships should be anonymized to prevent the compromise of individual privacy. To quantify the guarantee level of privacy-preserving mechanisms and mitigate users’ privacy concerns, numerous studies concerning network data de-anonymization have been carried out. However, most existing studies focus on single-view data, and privacy protection for multi-view data that is ubiquitous in the era of big data has not been yet extensively explored. In this study, we are interested in answering the following question: Are the traditional privacy protection methods still valid for the anonymization of multi-view data? In this study, we propose a Multi-View Low-Rank Coding (MVLRC) based network data de-anonymization framework to assess the vulnerability of privacy protection techniques by accurately reconstructing a large portion of the original data. Specifically, the framework assumes that in principle, the target and auxiliary networks have common structural patterns, and they can be modeled together to infer the hidden structure of the target network. The essential components of our work include the following: (1) a robust network representation model for structural pattern learning; (2) the network representation based multi-view modeling of target network and auxiliary network; (3) the inference of the anonymized links via target network reconstruction. Experimental results on synthetic networks and three real-world networks demonstrate that auxiliary networks can be utilized by malicious adversaries for privacy inference attacks. Thus, the privacy protection of multi-view network data needs more sophisticated anonymization techniques.

Copy Number Variation Pattern for Discriminating MACROD2 States of Colorectal Cancer Subtypes.

Copy number variation (CNV) is a common structural variation pattern of DNA, and it features a higher mutation rate than single-nucleotide polymorphisms (SNPs) and affects a larger fragment of genomes. CNV is related with the genesis of complex diseases and can thus be used as a strategy to identify novel cancer-predisposing markers or mechanisms. In particular, the frequent deletions of mono-ADP-ribosylhydrolase 2 (MACROD2) locus in human colorectal cancer (CRC) alters DNA repair and the sensitivity to DNA damage and results in chromosomal instability. The relationship between CNV and cancer has not been explained. In this study, on the basis of the genome variation profiling by the SNP array from 651 CRC primary tumors, we computationally analyzed the CNV data to select crucial SNP sites with the most relevance to three different states of MACROD2 (heterozygous deletion, homozygous deletion, and normal state), suggesting that these CNVs may play functional roles in CRC tumorigenesis. Our study can shed new insights into the genesis of cancer based on CNV, providing reference for clinical diagnosis, and treatment prognosis of CRC.