Representation Of Chemical Structures Research Articles

Basics in cheminformatics: representation of chemical structures 2

Basics in cheminformatics: representation of chemical structures 2

Basics in cheminformatics: representation of chemical structures 1

Basics in cheminformatics: representation of chemical structures 1

StructNet-DDI: Molecular Structure Characterization-Based ResNet for Prediction of Drug-Drug Interactions.

This study introduces a deep learning framework based on SMILES representations of chemical structures to predict drug-drug interactions (DDIs). The model extracts Morgan fingerprints and key molecular descriptors, transforming them into raw graphical features for input into a modified ResNet18 architecture. The deep residual network, enhanced with regularization techniques, efficiently addresses training issues such as gradient vanishing and exploding, resulting in superior predictive performance. Experimental results show that StructNet-DDI achieved an AUC of 99.7%, an accuracy of 94.4%, and an AUPR of 99.9%, demonstrating the model's effectiveness and reliability. These findings highlight that StructNet-DDI can effectively extract crucial features from molecular structures, offering a simple yet robust tool for DDI prediction.

MPOCSR: optical chemical structure recognition based on multi-path Vision Transformer

AbstractOptical chemical structure recognition (OCSR) is a fundamental and crucial task in the field of chemistry, which aims at transforming intricate chemical structure images into machine-readable formats. Current deep learning-based OCSR methods typically use image feature extractors to extract visual features and employ encoder-decoder architectures for chemical structure recognition. However, the performance of these methods is limited by their image feature extractors and the class imbalance of elements in chemical structure representation. This paper proposes MPOCSR (multi-path optical chemical structure recognition), which introduces the multi-path Vision Transformer (MPViT) and the class-balanced (CB) loss function to address these two challenges. MPOCSR uses MPViT as an image feature extractor, combining the advantages of convolutional neural networks and Vision Transformers. This strategy enables the provision of richer visual information for subsequent decoding processes. Furthermore, MPOCSR incorporates CB loss function to rebalance the loss weights among different categories. For training and validation of our method, we constructed a dataset that includes both Markush and non-Markush structures. Experimental results show that MPOCSR achieves an accuracy of 90.95% on the test set, surpassing other existing methods.

Indeks Padmakar-Ivan dan indeks Randic pada graf non-koprima dari grup bilangan bulat modulo

Graph theory, introduced by the Swiss mathematician Leonhard Euler in 1736, has played a pivotal role in solving real-world problems since its inception, notably exemplified by Euler's solution to the Konigsberg Bridge problem. Its applications extend to various domains, including scheduling, shortest path routing, and chemical structure representation. In chemistry, graphs are extensively used to depict molecular structures and chemical compounds, aiding in visualizing atomic connections and overall compound configurations. Topology indices, such as the Padmakar-Ivan (PI) and Randic indices, provide numerical values capturing chemical bonding relationships. Beyond chemical structures, these indices find applications in abstract algebraic graph representations. Recent research, exemplified by Husni et al.'s work on the harmonic and Gutman indices, explores these indices in coprime graphs of integer groups modulo prime power orders. Additionally, studies on non-coprime graphs of integer groups modulo reveal unique characteristics and invariants, shedding light on their structure. The non-coprime graph is a graph with two vertices said to be adjacent if the greatest common divisor (GCD) of their orders is not equal to one. This paper aims to investigate the topological indices, specifically the Padmakar-Ivan and Randic indices, in non-coprime graphs of integer groups modulo, adding depth to our understanding of their applicability and significance in abstract algebraic representations.
 Keywords: Graph theory, padmakar-ivan index, randic index, non-coprime graphsMSC2020: 05C09

Students’ Visualization Skills in Chemical Bonding through Three Dimension: A Study on Secondary School Chemistry Students

Visual cognitive representation of chemical structures and composition is achieved when proper visualization is made by chemistry students. It is in respect this that, this study measured the students’ visualization skills in chemical bonding through three dimension: a study on secondary school chemistry students. The study adopted quasi-experimental group design where both experimental (students taught using 3D visualization) and control (students taught using 2D visualization) groups participated in pre-test and post-test but only the experimental group received the intervention as 3D visualization. The population for this research comprises all science senior secondary school one (SSS1) students under Sokoto State Ministry of Science and Technology, Sokoto State, Nigeria. A sample of 118 students purposively selected were used. The instrument used for this research was Students’ Visualization Test in Chemical Bonding (SVTCB). The measured reliability of the instruments was 0.72. Independent sample t-test was used to analyse data collected. The findings implies that, 3D visualization enhances students’ visualization skills in the learning of chemistry abstract concept irrespective of gender. Keywords: Visual, Cognitive, Skills, Chemical Bonding, 3D

Using electrostatic potential maps as visual representations to promote better understanding of chemical bonding

Abstract Static visual representations (VRs) of chemical structures are necessary for an understanding of chemical bonding, a topic which continues to lead to learning difficulties and misconceptions for many students. The efficacy and problems associated with the use of VRs of chemical structures and chemical bonding in the form of electrostatic potential maps resulting from accurate quantum mechanical calculations are the subject of this study, which involved a sample of first year, second semester students, studying the elective course “Science Education” (N = 31). Students distinguished between nonpolar and polar covalent bonding, however, they encountered difficulties with concepts related to ionic bonding. Most students did not employ multistructural thinking (in the sense of the SOLO taxonomy), when providing explanations about the variation of bond polarity. Persistence of a covalent-ionic bond dichotomy was apparent, while for some, ions can be involved in both ionic and covalent bonding. Many students preferred to use their established high school knowledge. On a positive note, many students were clearly affected by the information provided by the colored VRs. Finally, the minimal experience of our students with these VRs leads us to believe that a more systematic and extensive coverage would be likely to produce improved outcomes.

Machine learning discovery of high-temperature polymers

SummaryTo formulate a machine learning (ML) model to establish the polymer's structure-property correlation for glass transition temperature , we collect a diverse set of nearly 13,000 real homopolymers from the largest polymer database, PoLyInfo. We train the deep neural network (DNN) model with 6,923 experimental values using Morgan fingerprint representations of chemical structures for these polymers. Interestingly, the trained DNN model can reasonably predict the unknown values of polymers with distinct molecular structures, in comparison with molecular dynamics simulations and experimental results. With the validated transferability and generalization ability, the ML model is utilized for high-throughput screening of nearly one million hypothetical polymers. We identify more than 65,000 promising candidates with > 200°C, which is 30 times more than existing known high-temperature polymers (∼2,000 from PoLyInfo). The discovery of this large number of promising candidates will be of significant interest in the development and design of high-temperature polymers.

Structure–property models of organic compounds based on molecular graphs with elements of the spatial structures of the molecules

Objectives. This article aims to describe, elaborate, and test a general algorithmic method for constructing the structure–property models for organic compounds.Results. The construction of the models is based on the statistical analysis of some sets of chemical structures of definite classes with known property values. These models have some forms of correlation equations. For the representation of chemical structures in this method, the special weighted molecular graphs (MGs) that reflect some peculiarities of the spatial structures of the corresponding molecules are used. The proposed method is realized in two steps. First, it is assumed that the required structure–property equation has a definite form and depends on several adjusted numerical parameters and two changeable functions of one variable. In this step, from some set of functions, the pair of functions that provide the best model is selected. In the second step, the best model (from the previous step) is modified. For this purpose, the classification of the vertices of MG by the chemical symbols of the corresponding atoms and their first-order environments is fulfilled. Further, the graph edges are classified according to the classes of the vertices which they connect. Furthermore, the numerical correction terms for the initial weights of the vertices and edges are introduced, and they improve the obtained model. The final result of the model-construction process is the equation of the definite form containing concrete numerical values of its parameters. Some examples of the application of the elaborated method for constructing the structure–property models for the concrete properties and classes of compounds are presented. The following classes of organic compounds and their physicochemical properties are considered: 1) the boiling point of alcohols, 2) the water solubility of alcohols, 3) the boiling point of sulfides, and 4) the retention indices of alkylphenols. The obtained results indicate the efficiency of the proposed approach and the significance of introducing the second step to the method.Conclusions. In this work, a general algorithmic and computerized method for constructing the structure–property models of organic compounds is suggested. Examples of the application of this method demonstrated its high efficiency. The method is suitable for any class of organic compounds and properties, which are quantitatively measured. Owing to its high efficiency, the structure–property models obtained by this approach can be employed to calculate the properties of chemical compounds for which experimental data are unavailable.

Machine Learning Discovery of High-Temperature Polymers

The design and development of high-temperature polymers have been an experimentally-driven and trial-and-error process guided by experience, intuition, and conceptual insights. However, such an Edisonian approach is often costly, slow, biased towards certain chemical space domains, and limited to relatively small-scale studies, which may easily miss promising compounds. To overcome this challenge, we formulate a data-driven machine-learning (ML) approach, integrated with high-fidelity molecular dynamics (MD) simulations, for quantitatively predicting polymer’s glass transition temperature (T_g) from its chemical structure and rapid screening of promising candidates for high-temperature polymers. Specifically, we collect a diverse set of nearly 13,000 real polymers from the largest polymer database, PoLyInfo. Among them, 6,923 experimental T_g values are available; while, the remaining 5,690 polymers do not have reported T_g values. We train the deep neural network (DNN) model with 6,923 experimental T_g values using Morgan fingerprint representations of chemical structures for these polymers. Interestingly, the trained DNN model can reasonably predict the unknown T_g values of polymers with distinct molecular structures, in comparison with MD simulations and experimental results. We find that this excellent transferability is attributed to the feature representation of Morgan fingerprints, which carry the chemical connectivity between neighboring repeating units in a polymer and the frequency of occurrence of different chemical substructures. With the validated ML model for high-throughput screening of nearly 1 million hypothetical polymers, we identify more than 65,000 promising candidates with T_g>200 oC, which is 30 times more than existing known high-temperature polymers (~2,000 from PoLyInfo). The discovery of this large number of promising candidates will be of significant interest in the development and design of high-temperature polymers. In short, our work demonstrates that ML is a powerful method for the prediction and rapid screening of high-temperature polymers, particularly with growing large sets of experimental and computational data for polymeric materials.

Adapting Interactive Interview Tasks to Remote Data Collection: Human Subjects Research That Requires Annotations and Manipulations of Chemical Structures During the COVID-19 Pandemic

In response to the Coronavirus-19 pandemic, higher education institutions quickly moved to emergency remote learning, and most human subjects research came to an immediate halt. Chemistry education researchers who wanted to continue data collection were confronted by many challenges, including research designs and data collection methods that had presumed local access to student participants, such as interactive tasks for students to complete during face-to-face interviews. In order to resume data collection remotely, interactive tasks needed to be translated to a virtual space. This communication reports adaptations made to an interview protocol to facilitate data collection in a remote setting with students as they interacted with and annotated multiple representations of chemical structures. The benefits and limitations to remotely capturing students’ drawings and interactions with multiple chemistry representations using the features available through a common video conferencing platform are discussed.

Graphical Representation Standards for Organic Chemistry Script: IUPAC Recommendation 2008

In this article, we will acquaint ourselves with IUPAC representation standards in organic chemistry. Most of the chemistry is expressed through graphical representations, and therefore, molecular structures shall be sketched carefully so that the drawing conveys exact and desired meaning. There are specific rules and regulations governing the graphical representations in chemistry. Less literature is available for this purpose, but the importance of the topic can be well understood after knowing that the division of chemical nomenclature and structure representation of IUPAC has published the Graphical Representation Standards for Chemical Structures in 2008 comprising 134 pages. The present article focuses on the etiquettes of structure drawings along with a summary in brief about the IUPAC 2008 recommendations in perspectives of organic chemistry.

Graph Convolutional Neural Networks as "General-Purpose" Property Predictors: The Universality and Limits of Applicability.

Nowadays the development of new functional materials/chemical compounds using machine learning (ML) techniques is a hot topic and includes several crucial steps, one of which is the choice of chemical structure representation. The classical approach of rigorous feature engineering in ML typically improves the performance of the predictive model, but at the same time, it narrows down the scope of applicability and decreases the physical interpretability of predicted results. In this study, we present graph convolutional neural networks (GCNNs) as an architecture that allows for successfully predicting the properties of compounds from diverse domains of chemical space, using a minimal set of meaningful descriptors. The applicability of GCNN models has been demonstrated by a wide range of chemical domain-specific properties. Their performance is comparable to state-of-the-art techniques; however, this architecture exempts from the need to carry out precise feature engineering.

Comparison of Data Representation Languages in the Structure–Activity Problem

This paper discusses two languages for representation of chemical structures and carries out their comparative analysis for the prediction of biological activity using an intelligent JSM-system. The comparison is carried out on three data arrays in terms of the group of parameters.

ADAPTIVE: leArning DAta-dePendenT, concIse molecular VEctors for fast, accurate metabolite identification from tandem mass spectra.

MotivationMetabolite identification is an important task in metabolomics to enhance the knowledge of biological systems. There have been a number of machine learning-based methods proposed for this task, which predict a chemical structure of a given spectrum through an intermediate (chemical structure) representation called molecular fingerprints. They usually have two steps: (i) predicting fingerprints from spectra; (ii) searching chemical compounds (in database) corresponding to the predicted fingerprints. Fingerprints are feature vectors, which are usually very large to cover all possible substructures and chemical properties, and therefore heavily redundant, in the sense of having many molecular (sub)structures irrelevant to the task, causing limited predictive performance and slow prediction.ResultsWe propose ADAPTIVE, which has two parts: learning two mappings (i) from structures to molecular vectors and (ii) from spectra to molecular vectors. The first part learns molecular vectors for metabolites from given data, to be consistent with both spectra and chemical structures of metabolites. In more detail, molecular vectors are generated by a model, being parameterized by a message passing neural network, and parameters are estimated by maximizing the correlation between molecular vectors and the corresponding spectra in terms of Hilbert-Schmidt Independence Criterion. Molecular vectors generated by this model are compact and importantly adaptive (specific) to both given data and task of metabolite identification. The second part uses input output kernel regression (IOKR), the current cutting-edge method of metabolite identification. We empirically confirmed the effectiveness of ADAPTIVE by using a benchmark data, where ADAPTIVE outperformed the original IOKR in both predictive performance and computational efficiency.Availability and implementationThe code will be accessed through http://www.bic.kyoto-u.ac.jp/pathway/tools/ADAPTIVE after the acceptance of this article.

Ligand-Based Virtual Screening Using Graph Edit Distance as Molecular Similarity Measure.

Extended reduced graphs provide summary representations of chemical structures using pharmacophore-type node descriptions to encode the relevant molecular properties. Commonly used similarity measures using reduced graphs convert these graphs into 2D vectors like fingerprints, before chemical comparisons are made. This study investigates the effectiveness of a graph-only driven molecular comparison by using extended reduced graphs along with graph edit distance methods for molecular similarity calculation as a tool for ligand-based virtual screening applications, which estimate the bioactivity of a chemical on the basis of the bioactivity of similar compounds. The results proved to be very stable and the graph editing distance method performed better than other methods previously used on reduced graphs. This is exemplified with six publicly available data sets: DUD-E, MUV, GLL&GDD, CAPST, NRLiSt BDB, and ULS-UDS. The screening and statistical tools available on the ligand-based virtual screening benchmarking platform and the RDKit were also used. In the experiments, our method performed better than other molecular similarity methods which use array representations in most cases. Overall, it is shown that extended reduced graphs along with graph edit distance is a combination of methods that has numerous applications and can identify bioactivity similarities in a structurally diverse group of molecules.

The use of social semiotic multimodality and joint action theory to describe teaching practices: two cases studies with experienced teachers

ABSTRACT The aim of this study is to better understand how teachers use various embodied semiotic modes – speech, gestures, gaze, and proxemics – when they present scientific content to the entire class. To analyse the multimodal practices of these teachers, we integrated the social semiotic theory of multimodality and the joint action theory in didactics. We use a case study methodology to analyse the video data and to describe the productions of two teachers from selected short episodes. The results show several similar elements of these teachers’ practices. In particular, their use of several coordinated modes to make a coherent meaning to complex aspects of scientific knowledge such as three- dimensional representations of chemical structures or the relation between an experiment and its model of light diffraction. Both teachers create signs by taking everyday things from their environment to re-signify them in a scientific way. Simultaneously they maintain contact with their students through their gaze and their position. In this way, they create or strengthen the classroom habits related to the explanation of knowledge.

Static to Dynamic: Kazan Chemists and the Transformation from Chemical Structure to Reaction Regiochemistry.

During the sesquicentennial of the first publication of Markovnikov's Rule, it is appropriate to re-visit the development of A. M. Butlerov's version of the theory of chemical structure, and its extension from molecular structure (static) to reaction regiochemistry (dynamic) by Butlerov's student (disciple!), V. V. Markovnikov.

Towards molecular design using 2D-molecular contour maps obtained from PLS regression coefficients

ABSTRACTThe multivariate image analysis descriptors used in quantitative structure-activity relationships are direct representations of chemical structures as they are simply numerical decodifications of pixels forming the 2D chemical images. These MDs have found great utility in the modeling of diverse properties of organic molecules. Given the multicollinearity and high dimensionality of the data matrices generated with the MIA-QSAR approach, modeling techniques that involve the projection of the data space onto orthogonal components e.g. Partial Least Squares (PLS) have been generally used. However, the chemical interpretation of the PLS-based MIA-QSAR models, in terms of the structural moieties affecting the modeled bioactivity has not been straightforward. This work describes the 2D-contour maps based on the PLS regression coefficients, as a means of assessing the relevance of single MIA predictors to the response variable, and thus allowing for the structural, electronic and physicochemical interpretation of the MIA-QSAR models. A sample study to demonstrate the utility of the 2D-contour maps to design novel drug-like molecules is performed using a dataset of some anti-HIV-1 2-amino-6-arylsulfonylbenzonitriles and derivatives, and the inferences obtained are consistent with other reports in the literature. In addition, the different schemes for encoding atomic properties in molecules are discussed and evaluated.

Was Duhem Justified in not Distinguishing Between Physical and Chemical Atomism?

Chemists in the late nineteenth century were apt to distinguish the theory of chemical structure they advocated as chemical, as opposed to physical, atomism. The failure on Duhem’s part to consider any such distinction in his critique of atomism might be taken to be a lacuna in his argument. Far from being a weakness in his stance, however, I argue that he had good systematic reasons for not taking such a distinction seriously.