Small Database Research Articles

Classification of COVID-19 Disease Based on Extra Tree Features Selection

The World Health Organization (WHO) has classified coronavirus as a global health emergency. Chest X-rays have been proven to be beneficial in both diagnosing and monitoring various lung diseases, including COVID-19. In this study, a COVID-19 disease detection framework is provided based on the methods used in machine learning and the Extra Tree algorithm to reduce the features extracted from the images. Using the Gray-Level Co-occurrence Matrix (GLCM) algorithm and the Extra Tree algorithm to choose the features of the work, a set of features is extracted from the images. This set of features is then put into the XGBoost algorithm to be classified. The proposed system was evaluated using two different sets of databases: the large database with 9544 images and the small database with 800 images. All image sizes were set to 300 x 300 pixels. The proposed system achieved a classification accuracy score of 90.04% using the large data set and 99.37% using the small set.

Designing potential lead compounds targeting aminoglycoside N (6′)-acetyltransferase in Serratia marcescens: A drug discovery strategy

Serratia marcescens is an opportunistic human pathogen that causes urinary tract infections, ocular lens infections, and respiratory tract infections. S. marcescens employs various defense mechanisms to evade antibiotics, one of which is mediated by aminoglycoside N-acetyltransferase (AAC). In this mechanism, the enzyme AAC facilitates the transfer and linkage of the acetyl moiety from the donor substrate acetyl-coenzyme A to specific positions on antibiotics. This modification alters the antibiotic's structure, leading to the inactivation of aminoglycoside antibiotics. In the current scenario, antibiotic resistance has become a global threat, and targeting the enzymes that mediate resistance is considered crucial to combat this issue. The study aimed to address the increasing global threat of antibiotic resistance in Serratia marcescens by targeting the aminoglycoside N-acetyltransferase (AAC (6′)) enzyme, which inactivates aminoglycoside antibiotics through acetylation. Due to the absence of experimental structure, we constructed a homology model of aminoglycoside N (6′)-acetyltransferase (AAC (6′)) of S. marcescens using the atomic structure of aminoglycoside N-acetyltransferase AAC (6′)-Ib (PDB ID: 1V0C) as a template. The stable architecture and integrity of the modelled AAC (6′) structure were analyzed through a 100 ns simulation. Structure-guided high-throughput screening of four small molecule databases (Binding, Life Chemicals, Zinc, and Toslab) resulted in the identification of potent inhibitors against AAC (6′). The hits obtained from screening were manually clustered, and the five hit molecules were shortlisted based on the docking score, which are observed in the range of −17.09 kcal/mol to −11.95 kcal/mol. These selected five molecules displayed acceptable pharmacological properties in ADME predictions. The binding free energy calculations, and molecular dynamics simulations of ligand bound AAC (6′) complexes represented higher affinity and stable binding. The selected molecules demonstrated stable binding with AAC (6′), indicating their strong potential to hamper the binding of aminoglycoside in the respective site. and thereby inhibit. This process mitigates enzyme mediated AAC (6′) activity on aminoglycosides and reverse the bactericidal function of aminoglycosides, and also this method could serve as a platform for the development of potential antimicrobials.

SMCVdb: a database of experimental cellular toxicity information for drug candidate molecules.

Many drug discovery exercises fail because small molecules that are effective inhibitors of target proteins exhibit high cellular toxicity. Early and effective assessment of toxicity and pharmacokinetics is essential to accelerate the drug discovery process. Conventional methods for toxicity profiling, including in vitro and in vivo assays, are laborious and resource-intensive. In response, we introduce the Small Molecule Cell Viability Database (SMCVdb), a comprehensive resource containing toxicity data for over 24 000 compounds obtained through high-content imaging (HCI). SMCVdb seamlessly integrates chemical descriptions and molecular weight data, offering researchers a holistic platform for toxicity data aiding compound prioritization and selection based on biological and economic considerations. Data collection for SMCVdb involved a systematic approach combining HCI toxicity profiling with chemical information and quality control measures ensured data accuracy and consistency. The user-friendly web interface of SMCVdb provides multiple search and filter options, allowing users to query the database based on compound name, molecular weight range, or viability percentage. SMCVdb empowers users to access toxicity profiles, molecular weights, compound names, and chemical descriptions, facilitating the exploration of relationships between compound properties and their effects on cell viability. In summary, the database provides experimentally derived cellular toxicity information for over 24 000 drug candidate molecules to academic researchers, and pharmaceutical companies. The SMCVdb will keep growing and will prove to be a pivotal resource to expedite research in drug discovery and compound evaluation. Database URL: http://smcvdb.rcb.ac.in:4321/.

Persons in Context. A Model to Represent Observations and Reconstructions of Historical Persons in Linked Data

Reconstruction of historical persons and family ties is the bread and butter of many researchers and genealogists. With the increasing digital availability of historical person records, the scope and depth of person reconstructions speaks to the imagination of researchers and genealogists. Yet, the lack of standardisation in the description of historical person data has hurt the interoperability and sustainability of both small and large databases. Persons in Context, or PiCo, presents a data model for historical person records within the Resource Description Framework (RDF). RDF or Linked Data is specifically designed for clear, unambiguous information exchange between multiple parties over the internet. We show how reuse of existing ontologies and concentric description are the building blocks of a flexible, straightforward, and stringent data model that emphasises provenance.

Dynamic Gradient Boosted Metaheuristic approach for efficient reinforced concrete structure optimization

Structural optimization remains a time-consuming and challenging task, particularly for reinforced concrete structures. Surrogate modeling, a subset of machine learning, has emerged as a promising approach to mitigate the computational burden of structural analysis. However, the use of surrogate modeling also introduces additional challenges, especially the risk of prediction errors due to a small training database. This paper introduces a novel framework, Dynamic Gradient Boosted Metaheuristic Method (DGBM), to address this challenge and enhance optimization efficiency. DGBM integrates Light Gradient Boosted Machine (LightGBM) as a surrogate model for structural analysis, along with a modified Symbiotic Organisms Search (SOS) algorithm, effectively exploring the search space to discover optimal solutions. Its innovative features include LightGBM screening, dynamic training, design modifications, adaptive correction mechanisms, and integration with ETABS. The current study aims to leverage the strengths of machine learning and metaheuristics to significantly reduce the computational cost associated with traditional optimization methods while minimizing the impact of prediction errors. These enhancements refine and expedite the optimization process, making it more applicable to real-world design scenarios. The effectiveness of DGBM is evaluated through case studies on practical reinforced concrete structures, the proposed framework consistently achieves the target cost and delivers results 42.5 % faster in case 1, 57.1 % faster in case 2, and 25 % faster in case 3 compared to the next best method. This advantage translates to faster design iterations and increased efficiency for engineers, empowering them to explore a broader range of design options and achieve optimal solutions for complex structures in a shorter timeframe.

1798P Relapsed and refractory systemic therapy real-world outcomes in the Canadian small cell lung cancer database (CASCADE)

1798P Relapsed and refractory systemic therapy real-world outcomes in the Canadian small cell lung cancer database (CASCADE)

1793P Treatment and outcomes of limited disease small cell lung cancer (LD-SCLC) in the Canadian small cell lung cancer database (CASCADE)

1793P Treatment and outcomes of limited disease small cell lung cancer (LD-SCLC) in the Canadian small cell lung cancer database (CASCADE)

A Methodological Approach to Assembly Time Standard Estimation Based on Incomplete Characteristics of the Production Process and Using Small Databases

The problem solved in this article concerns assembly planning, which is time-consuming, but crucial in the development of mechanical products. At the product design stage there is no complete information about the manufacturing process, so it is necessary to develop an approach to help process the uncertain and incomplete information. In order to compare different product variants, the assembly time standard has to be estimated on the basis of the incomplete product and production process characteristics. This paper presents a method for estimating the assembly time standard using time classes, decision tree and evidence theory.

Maximum radial pattern matching for minimum star map identification

This paper proposes an all-sky star map identification algorithm that can simultaneously achieve high identification probability, low algorithm complexity, and small databases for well photometric and intrinsic parameters-calibrated star sensors. The proposed algorithm includes three main steps. First, a binary radial pattern table is constructed offline. Then, the maximum value matching of the radial pattern is performed between the star spots and the guide stars, and the star pairs (i.e., the minimum star map) after radial pattern matching undergo a coarse matching through angular distance cross-validation. Finally, a reference star map is designed based on the identified star pairs, and the matching of all the star spots in the field of view is realized. Simulation and analysis results show that the database required by the proposed algorithm for 5,000 guide stars is not larger than 200 KB. Also, when false and missing star spots account for 50% of all guide stars and the star spot extraction error is 0.5 pixel (the corresponding pointing error is 26″), the average star map identification time of the proposed algorithm is less than 2 ms, and its identification probability is higher than 98%. The results demonstrate that the proposed algorithm performs better than similar algorithms.

A Small Database with an Adaptive Data Selection Method for Solder Joint Fatigue Life Prediction in Advanced Packaging.

There has always been high interest in predicting the solder joint fatigue life in advanced packaging with high accuracy and efficiency. Artificial Intelligence Plus (AI+) is becoming increasingly popular as computational facilities continue to develop. This study will introduce machine learning (a core component of AI). With machine learning, metamodels that approximate the attributes of systems or functions are created to predict the fatigue life of advanced packaging. However, the prediction ability is highly dependent on the size and distribution of the training data. Increasing the amount of training data is the most intuitive approach to improve prediction performance, but this implies a higher computational cost. In this research, the adaptive sampling methods are applied to build the machine learning model with a small dataset sampled from an existing database. The performance of the model will be visualized using predefined criteria. Moreover, ensemble learning can be used to improve the performance of AI models after they have been fully trained.

Predicting the Maximum Displacements of Structures during an Earthquake based on Chaos Theory

The maximum displacements are an important index for the structural health monitoring (SHM) after an earthquake. However, the measurement of displacements is not easy compared to that of accelerations, and nonlinear structural identification is problematic. This study estimates displacements using input and output accelerations with a small database without assuming any physical models. A nonlinear-dynamics prediction algorithm based on chaos theory is proposed combining Takens’s embedding theorem and nearest-neighbor method. Using 12 ground motions in simulation and three motions in experimental cases, the maximum displacements are predicted with 8 % to 25 % mean absolute errors (MAEs). A practical method is developed for robustly and accurately evaluating displacements.

Reliability analysis of various modeling techniques for the prediction of axial strain of FRP-confined concrete

PurposeThe external confinement provided by the fiber-reinforced polymer (FRP) sheets leads to an improvement in the axial compressive strength (CS) and strain of reinforced concrete structural members. Many studies have proposed analytical models to predict the axial CS of concrete structural members, but the predictions for the axial compressive strain still need more investigation because the previous strain models are not accurate enough. Moreover, the previous strain models were proposed using small and noisy databases using simple modeling techniques. Therefore, a rigorous approach is needed to propose a more accurate strain model and compare its predictions with the previous models.Design/methodology/approachThe present work has endeavored to propose strain models for FRP-confined concrete members using three different techniques: analytical modeling, artificial neural network (ANN) modeling and finite element analysis (FEA) modeling based on a large database consisting of 570 sample points.FindingsThe assessment of the previous models using some statistical parameters revealed that the estimates of the newly recommended models were more accurate than the previous models. The estimates of the new models were validated using the experimental outcomes of compressive members confined with carbon-fiber-reinforced polymer (CFRP) wraps. The nonlinear FEA of the tested samples was performed using ABAQUS, and its estimates were equated with the calculations of the analytical and ANN models. The relative investigation of the estimates solidly substantiates the accuracy and applicability of the recommended analytical, ANN and FEA models for predicting the axial strain of CFRP-confined concrete compression members.Originality/valueThe research introduces innovative methods for understanding FRP confinement in concrete, presenting new models to estimate axial compressive strains. Utilizing a database of 570 experimental samples, the study employs ANNs and regression analysis to develop these models. Existing models for FRP-confined concrete's axial strains are also assessed using this database. Validation involves testing 18 cylindrical specimens confined with CFRP wraps and FE simulations using a concrete-damaged plastic (CDP) model. A comprehensive comparative analysis compares experimental results with estimates from ANNs, analytical and finite element models (FEMs), offering valuable insights and predictive tools for FRP confinement in concrete.

Metabolomic profiling analysis reveals the benefits of ginseng berry intake on mitochondrial function and glucose metabolism in the liver of obese mice.

Ginseng berry (GB) has previously been demonstrated to improve systemic insulin resistance and regulate hepatic glucose metabolism and steatosis in mice with diet-induced obesity (DIO). In this study, the role of GB in metabolism was assessed using metabolomics analysis on the total liver metabolites of DIO mice. Metabolomic profiling was performed using capillary electrophoresis time-of-flight mass spectrometry (CE-TOF/MS) of liver tissue from mice on a 12-wk normal chow diet (NC), high-fat diet (HFD), and HFD supplemented with 0.1% GB (HFD + GB). The detected metabolites, its pathways, and functions were analyzed through partial least square discriminant analysis (PLS-DA), the small molecular pathway database (SMPDB), and MetaboAnalyst 5.0. The liver metabolite profiles of NC, HFD, and GB-fed mice (HFD + GB) were highly compartmentalized. Metabolites involved in major liver functions, such as mitochondrial function, gluconeogenesis/glycolysis, fatty acid metabolism, and primary bile acid biosynthesis, showed differences after GB intake. The metabolites that showed significant correlations with fasting blood glucose (FBG), insulin, and homeostatic model assessment for insulin resistance (HOMA-IR) were highly associated with mitochondrial membrane function, energy homeostasis, and glucose metabolism. Ginseng berry intake increased the levels of metabolites involved in mitochondrial membrane function, decreased the levels of metabolites related to glucose metabolism, and was highly correlated with metabolic phenotypes. This study demonstrated that long-term intake of GB changed the metabolite of hepatosteatotic livers in DIO mice, normalizing global liver metabolites involved in mitochondrial function and glucose metabolism and indicating the potential mechanism of GB in ameliorating hyperglycemia in DIO mice.

Transfer learning across different chemical domains: virtual screening of organic materials with deep learning models pretrained on small molecule and chemical reaction data

Machine learning is becoming a preferred method for the virtual screening of organic materials due to its cost-effectiveness over traditional computationally demanding techniques. However, the scarcity of labeled data for organic materials poses a significant challenge for training advanced machine learning models. This study showcases the potential of utilizing databases of drug-like small molecules and chemical reactions to pretrain the BERT model, enhancing its performance in the virtual screening of organic materials. By fine-tuning the BERT models with data from five virtual screening tasks, the version pretrained with the USPTO–SMILES dataset achieved R2 scores exceeding 0.94 for three tasks and over 0.81 for two others. This performance surpasses that of models pretrained on the small molecule or organic materials databases and outperforms three traditional machine learning models trained directly on virtual screening data. The success of the USPTO–SMILES pretrained BERT model can be attributed to the diverse array of organic building blocks in the USPTO database, offering a broader exploration of the chemical space. The study further suggests that accessing a reaction database with a wider range of reactions than the USPTO could further enhance model performance. Overall, this research validates the feasibility of applying transfer learning across different chemical domains for the efficient virtual screening of organic materials.Scientific contributionThis study verifies the feasibility of applying transfer learning to large language models in different chemical fields to help organic materials perform virtual screening. Through the comparison of transfer learning from different chemical fields to a variety of organic material molecules, the high precision virtual screening of organic materials is realized.

Small dataset machine-learning approach for efficient design space exploration: engineering ZnTe-based high-entropy alloys for water splitting

Aiming toward a sustainable energy era, the design of efficient photocatalysts for water splitting by engineering their band properties has been actively studied. One promising avenue for the band engineering of active photocatalysts is the use of solid-solution alloying. However, the enormous possible configurations of multicomponent alloys hinders the experimental screening of this multidimensional material space, providing an opportunity for machine learning (ML) approaches to help accelerate the discovery of new multicomponent alloy materials. A conventional prerequisite for ML approaches is a large database of accurate material properties, which may require exhaustive computational and/or experimental resources. This study demonstrates that the screening of solid-solution alloys (up to hexanary systems) can be performed using a small database to minimize (and optimize) the number of high-level computational calculations. Specifically, we use ZnTe-based alloys as a prototypical example and employ a secure independent screening and sparsifing operator with the recently developed agreement method (α-method). Furthermore, we discuss and propose design routes to determine the optimal solid-solution ZnTe-based alloys for photoassisted water-splitting reactions.

Small Representative Databases for Testing and Validating Density Functionals and Other Electronic Structure Methods.

Broad and diverse sets of accurate data provide useful metrics for assessing the performance of new theoretical methods. However, assessing methods against large databases can be an arduous task. Here, we present 17 representative energetic databases, defined as small databases whose errors and error spreads are representative of larger databases and which therefore can serve as efficient benchmarks for developing and testing electronic structure methods and density functionals. In 15 cases, the representative databases have 6 entries while being representative of larger databases with 14-107 entries, and in the other two cases, they have 14 entries while being representative of larger databases with 418-455 entries. The mean unsigned error (MUE) of 100 electronic structure methods on a given representative database is typically within about 8% of the MUE on its parent database, and the root-mean-square error (RMSE) is typically within about 11% of the RMSE on the parent database. Thus, the representative databases are quite successful in indicating accuracy while maintaining good diversity. The databases include both main-group and transition-metal compounds and reactions, and they include bond energies, reaction energies, barrier heights, noncovalent interactions, ionization potentials, and absolute energies.

Generalisation capabilities of machine-learning algorithms for the detection of the subthalamic nucleus in micro-electrode recordings.

Micro-electrode recordings (MERs) are a key intra-operative modality used during deep brain stimulation (DBS) electrode implantation, which allow for a trained neurophysiologist to infer the anatomy in which the electrode is placed. As DBS targets are small, such inference is necessary to confirm that the electrode is correctly positioned. Recently, machine learning techniques have been used to augment the neurophysiologist's capability. The goal of this paper is to investigate the generalisability of these methods with respect to different clinical centres and training paradigms. Five deep learning algorithms for binary classification of MER signals have been implemented. Three databases from two different clinical centres have also been collected with differing size, acquisition hardware, and annotation protocol. Each algorithm has initially been trained on the largest database, then either directly tested or fine-tuned on the smaller databases in order to estimate their generalisability. As a reference, they have also been trained from scratch on the smaller databases as well in order to estimate the effect of the differing database sizes and annotation systems. Each network shows significantly reduced performance (on the order of a 6.5% to 16.0% reduction in balanced accuracy) when applied out-of-distribution. This reduction can be ameliorated through fine-tuning the network on the new database through transfer learning. Although, even for these small databases, it appears that retraining from scratch may still offer equivalent performance as fine-tuning with transfer learning. However, this is at the expense of significantly longer training times. Generalisability is an important criterion for the success of machine learning algorithms in clinic. We have demonstrated that a variety of recent machine learning algorithms for MER classification are negatively affected by domain shift, but that this can be quickly ameliorated through simple transfer learning procedures that can be readily performed for new centres.

A domain adaptation‐based convolutional neural network incorporating data augmentation for power system dynamic security assessment

AbstractRecently, deep learning (DL) based dynamic security assessment (DSA) methods have been very successful. However, although a DSA model can be trained well for a specific topology, it often does not perform well for other topologies. Since the topology in real‐world power systems is frequently changing, the performance reduction of DL‐based DSA methods is very serious, which is a challenging and urgent problem. This paper proposes a novel DSA method based on a convolutional neural network (CNN) to solve this problem. In the proposed method, a strong yet simple domain adaptation approach named adaptive batch normalization (AdaBN) is used, which significantly enhances the extensibility and generalizability of the DSA model when the topology changes and eliminates the need to train a large number of models. This approach achieves a deep adaptation effect by modulating the statistics from the source domain to the target domain in all batch normalization layers across the model. Unlike other domain adaptation methods, this method is parameter‐free, requires no additional components, and has advanced performance despite its simplicity. In addition, this paper introduces TGAN‐based data augmentation to deal with the difficulty of costly data collection and labelling. This data augmentation makes the proposed model applicable to small databases. The test results of the proposed method on IEEE 39‐bus and IEEE 118‐bus systems show that this method can evaluate system dynamic security during topology changes and in the face of data noise with high accuracy.

Ensuring Melt Track Width Consistency and Crack-Free Conditions Using Interpass-Temperature-Dependent Process Parameters for Wire-Arc-Directed Energy-Deposited Inconel 718

Melt track width can vary in a wire-arc-directed energy-deposited material (DED) using a constant set of process parameters, leading to a lower-quality build. In this work, a novel framework is proposed that uses the data from the process parameter development stage to create optimized process parameters for a target layer width at different interpass temperatures without hot cracking. Inconel 718 is used as the model material since it is known to suffer from hot cracking during DED processing. In the proposed framework, a process window containing a few sets of process parameters (torch travel speed and wire feed rate) is established for crack-free deposition of Inconel 718, and these parameters are used to create a small database. A linear regression model is then employed to generate interpass-temperature-specific optimized process parameters for a target melt track width. The results demonstrate that the proposed approach can reduce the melt track width variation in the deposited walls from 12% to 3% error on average under different printing conditions. It also demonstrates that interpass temperature (IPT) can be used as a controlled variable and the optimized process parameters as initial values when applying control techniques to the process.

Prediction of micropollutant degradation kinetic constant by ultrasonic using machine learning

A prediction model based on XGBoost is proposed for ultrasonic degradation of micropollutants' kinetic constants. After parameter optimization through iteration, the model achieves Evaluation metrics with R2 and SMAPE reaching 0.99 and 2.06%, respectively. The impact of design parameters on predicting kinetic constants for ultrasound degradation of trace pollutants was assessed using Shapley additive explanations (SHAP). Results indicate that power density and frequency significantly impact the predictive performance. The database was sorted based on power density and frequency values. Subsequently, 800 raw data were split into small databases of 200 each. After confirming that reducing the database size doesn't affect prediction accuracy, ultrasound degradation experiments were conducted for five pollutants, yielding experimental data. A small database with experimental conditions within the numerical range was selected. Data meeting both feature conditions were filtered, resulting in an optimized 60-data group. After incorporating experimental data, a model was trained for prediction. Degradation kinetic constants for experiments (kE) were compared with predicted constants (for 800 data-based model: kP-800 and for 60 data-based model: kP-60). Results showed ibuprofen, bisphenol A, carbamazepine, and 17β-Estradiol performed better on the 60-data group (kP-60/kE: 1.00, 0.99, 1.00, 1.00), while caffeine suited the model trained on the 800-data group (kP-800/kE: 1.02).