Structural Database Research Articles

Experiential Learning about Database Design in the MLIS Curriculum: Software and Instructional Design for Threshold Knowledge

With almost all MLIS and similar degree programs having a required course covering database design, it is critical to develop curriculum that engages students in learning essential concepts and how to implement designs. We report on a project that explored how essential knowledge about database design can be taught effectively through experiential learning activities involving collaborative groupwork and using database software to build small databases and search forms. The focus is an existing required introductory course on information retrieval system design, in which students learn about database design concepts and implement their designs in browser-based software, building the database structure, content, and search forms. They learn about critical concepts in database design, including eliciting user needs and requirements, prototyping, teamwork, and iterative testing and evaluation. The project had two aims: to select new database software to use as a learning tool in the course and to redesign course assignments so that learning experiences would be enhanced. The project led to important findings about the student learning experiences, to improvements in the curriculum, and to recommendations for educators involved in similar courses.

Decentralized clinical trials are better for the participants and for the planet: the case study of a double-blind randomized controlled trial in Singapore (PROMOTE study)

IntroductionNovel technologies have enabled the decentralization of many aspects of clinical trials, but little research has been done on the impact of these changes on the participant experience, trial operations, or the environment.MethodsA fully decentralized clinical trial conducted in Singapore is used as a case study to evaluate the operational outcomes, environmental impact (via life cycle assessment), and participants experience (qualitative interviews) of the decentralized model compared to a traditional study with in-person visits.ResultsThe decentralized study achieved high participant retention rates (97%) and high completion rates for clinical data, even for biological samples. Participants found the decentralized model to be convenient and safe, especially during the pandemic. Moreover, the decentralized model was found to be more environmentally friendly and less detrimental to human health compared to traditional face-to-face clinical trials, primarily by reducing participants’ use of cars for site visits.DiscussionWhile this study focused on the environmental impact, it is important to consider other factors such as participant safety, convenience, and data quality when evaluating the suitability of a decentralized clinical trial approach. Careful planning of data flow, database structure, and data protection measures is essential. This study contributes to improving the environmental footprint of clinical trials. Environmental sustainability should be among the factors that are evaluated when selecting trial models. Decentralized and hybrid clinical trials offer efficiency, effectiveness, and environmental benefits. Further research and adoption of these approaches are encouraged.

CPconf_score: A Deep Learning Free Energy Function Trained Using Molecular Dynamics Data for Cyclic Peptides.

Accurate structural feature characterization of cyclic peptides (CPs), especially those with less than 10 residues and cis-peptide bonds, is challenging but important for the rational design of bioactive peptides. In this study, we performed high-temperature molecular dynamics (high-T MD) simulations on 250 CPs with random sequences and applied the point-adaptive k-nearest neighbors (PAk) method to estimate the free energies of millions of sampled conformations. Using this data set, we trained a SchNet-based deep learning model, termed CPconf_score, to predict the conformational free energies of CPs. We tested CPconf_score to identify near-native conformations from MD-sampled conformations of 50 CPs from the Cambridge Structural Database. Our method achieved accurate predictions for 41 out of 50 CPs with a backbone RMSD of less than 1.0 Å compared to crystal structures. In comparison, other advanced CP structure prediction tools, such as HighFold and Rosetta, successfully predicted 12 and 19 CPs, respectively.

Force-Activated Spin-Crossover in Fe2+ and Co2+ Transition Metal Mechanophores.

Transition metal mechanophores exhibiting force-activated spin-crossover are attractive design targets, yet large-scale discovery of them has not been pursued due in large part to the time-consuming nature of trial-and-error experiments. Instead, we leverage density functional theory (DFT) and external force explicitly included (EFEI) modeling to study a set of 395 feasible Fe2+ and Co2+ mechanophore candidates with tridentate ligands that we curate from the Cambridge Structural Database. Among nitrogen-coordinating low-spin complexes, we observe the prevalence of spin crossover at moderate force, and we identify 155 Fe2+ and Co2+ spin-crossover mechanophores and derive their threshold force for low-spin to high-spin transition (FSCO). The calculations reveal strong correlations of FSCO with spin-splitting energies and coordination bond lengths, facilitating rapid prediction of FSCO using force-free DFT calculations. Then, among all Fe2+ and Co2+ spin-crossover mechanophores, we further identity 11 mechanophores that combine labile spin-crossover and good mechanical robustness that are thus predicted to be the most versatile for force-probing applications. We discover two classes of mer-symmetric complexes comprising specific heteroaromatic rings within extended π-conjugation that give rise to Fe2+ mechanophores with these characteristics. We expect the set of spin-crossover mechanophores, the design principles, and the computational approach to be useful in guiding the high-throughput discovery of transition metal mechanophores with diverse functionalities and broad applications, including mechanically activated catalysis.

Data-driven prediction of chemically relevant compositions in multi-component systems using tensor embeddings

The discovery of novel materials is crucial for developing new functional materials. This study introduces a predictive model designed to forecast complex multi-component oxide compositions, leveraging data derived from simpler pseudo-binary systems. By applying tensor decomposition and machine learning techniques, we transformed pseudo-binary oxide compositions from the Inorganic Crystal Structure Database (ICSD) into tensor representations, capturing key chemical trends such as oxidation states and periodic positions. Tucker decomposition was utilized to extract tensor embeddings, which were used to train a Random Forest classifier. The model successfully predicted the existence probabilities of pseudo-ternary and quaternary oxides, with 84% and 52% of ICSD-registered compositions, respectively, achieving high scores. Our approach highlights the potential of leveraging simpler oxide data to predict more complex compositions, suggesting broader applicability to other material systems such as sulfides and nitrides.

AlphaFold-guided molecular replacement for solving challenging crystal structures.

Molecular replacement (MR) is highly effective for biomolecular crystal structure determination, increasingly so as the database of known structures has increased. For candidates without recognizable similarity to known structures, however, crystal structure analyses have nearly always required experiments for de novo phase evaluation. Now, with the unprecedented accuracy of AlphaFold predictions of protein structures from amino-acid sequences, an appreciable expansion of the reach of MR for proteins is realized. Here, we sought to automate an AlphaFold-guided MR procedure that tailors predictions to the MR problem at hand. We first optimized the reliability cutoff parameters for residue inclusion as tested in application to a previously MR-intractable problem. We then examined cases where AlphaFold by default predicts a conformation alternative to that of the candidate structure, devising tests for MR solution either from domain-specific predictions or from predictions based on diverse sequence subclusters. We tested subclustering procedures on an enzyme system that entails multiple MR-challenging conformations. The overall process as implemented in Phenix automatically surveys a succession of trials of increasing computational complexity until an MR solution is found or the options are exhausted. Validated MR solutions were found for 92% of one set of 158 challenging problems from the PDB and 93% of those from a second set of 215 challenges. Thus, many crystal structure analyses that previously required experimental phase evaluation can now be solved by AlphaFold-guided MR. In effect, this and related MR approaches are de novo phasing methods.

The coordination chemistry of p-tert-butylcalix[8]arene with transition and lanthanide metal ions.

A survey of the literature and the Cambridge Structural Database reveals thirty nine p-tert-butylcalix[8]arene-supported transition (3d, limited to V-Cu) and lanthanide metal complexes ranging in nuclearity from one to eighteen, twenty of which are homometallic and nineteen of which are heterometallic. We provide a review of the coordination chemistry of these complexes, including our own work in the area. We also provide our thoughts and perspectives on the common structural themes observed, identify gaps in knowledge and evaluate how that may inform and direct future synthetic efforts. What is clear is that the coordination chemistry of H8TBC[8] with paramagnetic metal ions remains hugely underexplored.

Evidence of strong O-H⋯C interactions involving apical pyramidane carbon atoms as hydrogen atom acceptors.

Using high-level quantum chemical calculations, we predicted a strong O-H⋯C interaction between the apical carbon atoms of pyramidane and its derivatives and water molecules. Analysis of calculated electrostatic potential maps showed that there are areas of strong negative potential above apical carbon atoms in all studied structures. The results of quantum chemical calculations showed that the O-H⋯C interaction between the hydrogen atom of water and the apical carbon atom of pyramidane derivatives with four -CH3 substituents is unexpectedly strong, ΔECCSD(T)/CBS = -7.43 kcal mol-1. The strong hydrogen bonds were also predicted in the case of unsubstituted pyramidane (ΔECCSD(T)/CBS = -6.41 kcal mol-1) and pyramidane with four -OH substituents (ΔECCSD(T)/CBS = -5.87 kcal mol-1). Although there are not many crystal structures of pyramidane-like molecules, we extracted examples of pyramidal-shaped molecules with apical carbon atoms from the Cambridge Structural Database and analyzed their hydrogen-bonding patterns. Analysis of crystal structures confirmed the existence of short non-covalent contacts between apical carbon atoms and neighboring hydrogen atoms.

The Iconic α-Helix: From Pauling to the Present.

The protein folding problem dates back to Pauling's insights almost a century ago, but the first venture into actual protein structure was the Pauling-Corey-Brandson α-helix in 1951, a proposed model that was confirmed almost immediately using X-ray crystallography. Many subsequent efforts to predict protein helices from the amino acid sequence met with only partial success, as discussed here. Surprisingly, in 2021, these efforts were superseded by deep-learning artificial intelligence, especially AlphaFold2, a machine learning program based on neural nets. This approach can predict most protein structures successfully at or near atomic resolution. Deservedly, deep-learning artificial intelligence was named Science magazine's 2021 "breakthrough of the year." Today, ~200 million predicted protein structures can be downloaded from the AlphaFold2 Protein Structure Database. Deep learning represents a deep conundrum because these successfully predicted macromolecular structures are based on methods that are completely devoid of a hypothesis or of any physical chemistry. Perhaps we are now poised to transcend five centuries of reductive science.

The Evolving T Cell Receptor Recognition Code: The Rules Are More Like Guidelines.

αβ T cell receptor (TCR) recognition of peptide-MHC complexes lies at the core of adaptive immunity, balancing specificity and cross-reactivity to facilitate effective antigen discrimination. Early structural studies established basic frameworks helpful for understanding and contextualizing TCR recognition and features such as peptide specificity and MHC restriction. However, the growing TCR structural database and studies launched from structural work continue to reveal exceptions to common assumptions and simplifications derived from earlier work. Here we explore our evolving understanding of TCR recognition, illustrating how structural and biophysical investigations regularly uncover complex phenomena that push against paradigms and expand our understanding of how TCRs bind to and discriminate between peptide/MHC complexes. We discuss the implications of these findings for basic, translational, and predictive immunology, including the challenges in accounting for the inherent adaptability, flexibility, and occasional biophysical sloppiness that characterize TCR recognition.

Leveraging natural language processing to curate the tmCAT, tmPHOTO, tmBIO, and tmSCO datasets of functional transition metal complexes.

The breadth of transition metal chemical space covered by databases such as the Cambridge Structural Database and the derived computational database tmQM is not conducive to application-specific modeling and the development of structure-property relationships. Here, we employ both supervised and unsupervised natural language processing (NLP) techniques to link experimentally synthesized compounds in the tmQM database to their respective applications. Leveraging NLP models, we curate four distinct datasets: tmCAT for catalysis, tmPHOTO for photophysical activity, tmBIO for biological relevance, and tmSCO for magnetism. Analyzing the chemical substructures within each dataset reveals common chemical motifs in each of the designated applications. We then use these common chemical structures to augment our initial datasets for each application, yielding a total of 21 631 compounds in tmCAT, 4599 in tmPHOTO, 2782 in tmBIO, and 983 in tmSCO. These datasets are expected to accelerate the more targeted computational screening and development of refined structure-property relationships with machine learning.

DPAM-AI: a domain parser for AlphaFold models powered by artificial intelligence.

Due to the breakthrough in protein structure prediction by AlphaFold, the scientific community has access to 200 million predicted protein structures with near-atomic accuracy from the AlphaFold protein structure DataBase (AFDB), covering nearly the entire protein universe. Segmenting these models into domains and classifying them into an evolutionary hierarchy hold tremendous potential for unraveling essential insights into protein function. We introduce DPAM-AI, a Domain Parser for AlphaFold Models based on Artificial Intelligence. DPAM-AI utilizes a convolutional neural network trained with previously classified domains in the Evolutionary Classification Of protein Domains (ECOD) database. DPAM-AI integrates inter-residue distances, predicted aligned errors, and sequence and structural alignments to previously classified domains detected via sequence (HHsuite) and structural (Dali) similarity searches. DPAM-AI has demonstrated its power through rigorous tests, excelling in several benchmark sets compared to its predecessor, DPAM, and other recently published domain parsers, Merizo and Chainsaw. We applied DPAM-AI to representative AFDB models for proteins classified in Pfam. We obtained representative 3D structures for 18487 (89%) of the 20795 Pfam families. The remaining families either (i) belong to viral proteins that were excluded from AFDB or (ii) do not adopt globular 3D structures. Our structure-aware domain delineation uncovered a considerable fraction (15%) of Pfam domains containing multiple structural and evolutionary units and refined the boundaries for over half. Pfam and corresponding DPAM-AI domains are at http://prodata.swmed.edu/DPAM-pfam/. Our code is deposited at https://github.com/Jsauce5p/DPAM/tree/dpam_ai, and updates will be released through https://github.com/CongLabCode/DPAM.

Optimizing Hybrid Cloud Database Architecture: Integrating SQL Server and MongoDB in Azure Environments

This study explores the optimization of hybrid cloud database architectures by integrating SQL Server and MongoDB within Azure environments, focusing on the specific benefits and challenges of such an approach. Hybrid cloud architectures are increasingly adopted by organizations seeking flexibility, scalability, and improved data management. By examining the integration of SQL Server’s relational database structure with MongoDB’s NoSQL capabilities, this study aims to demonstrate how businesses can achieve efficient data synchronization, enhanced security, and improved performance within the Azure cloud platform. Through a comprehensive review of existing literature, analysis of case studies, and discussion of best practices, this paper provides actionable insights for organizations considering hybrid database solutions. Key findings indicate that optimizing hybrid cloud architectures with SQL Server and MongoDB can offer substantial benefits for diverse applications, but also presents unique challenges that must be managed through strategic planning and robust architecture design. This research draws on studies and publications from U.S. journals to provide a contemporary perspective on hybrid cloud database optimization.

The ab initio non-crystalline structure database: empowering machine learning to decode diffusivity

Non-crystalline materials exhibit unique properties that make them suitable for various applications in science and technology, ranging from optical and electronic devices and solid-state batteries to protective coatings. However, data-driven exploration and design of non-crystalline materials is hampered by the absence of a comprehensive database covering a broad chemical space. In this work, we present the largest computed non-crystalline structure database to date, generated from systematic and accurate ab initio molecular dynamics (AIMD) calculations. We also show how the database can be used in simple machine-learning models to connect properties to composition and structure, here specifically targeting ionic conductivity. These models predict the Li-ion diffusivity with speed and accuracy, offering a cost-effective alternative to expensive density functional theory (DFT) calculations. Furthermore, the process of computational quenching non-crystalline structures provides a unique sampling of out-of-equilibrium structures, energies, and force landscape, and we anticipate that the corresponding trajectories will inform future work in universal machine learning potentials, impacting design beyond that of non-crystalline materials. In addition, combining diffusion trajectories from our dataset with models that predict liquidus viscosity and melting temperature could be utilized to develop models for predicting glass-forming ability.

Determining structures of RNA conformers using AFM and deep neural networks.

Much of the human genome is transcribed into RNAs1, many of which contain structural elements that are important for their function. Such RNA molecules-including those that are structured and well-folded2-are conformationally heterogeneous and flexible, which is a prerequisite for function3,4, but this limits the applicability of methods such as NMR, crystallography and cryo-electron microscopy for structure elucidation. Moreover, owing to the lack of a large RNA structure database, and no clear correlation between sequence and structure, approaches such as AlphaFold5 for protein structure prediction do not apply to RNA. Therefore, determining the structures of heterogeneous RNAs remains an unmet challenge. Here we report holistic RNA structure determination method using atomic force microscopy, unsupervised machine learning and deep neural networks (HORNET), a novel method for determining three-dimensional topological structures of RNA using atomic force microscopy images of individual molecules in solution. Owing to the high signal-to-noise ratio of atomic force microscopy, this method is ideal for capturing structures of large RNA molecules in distinct conformations. In addition to six benchmark cases, we demonstrate the utility of HORNET by determining multiple heterogeneous structures of RNase P RNA and the HIV-1 Rev response element (RRE) RNA. Thus, our method addresses one of the major challenges in determining heterogeneous structures of large and flexible RNA molecules, and contributes to the fundamental understanding of RNA structural biology.

Data Governance and Distribution of Biobank: A Case from a Chinese Cancer Hospital.

Objectives: To facilitate the regionalization, specialization, and digitization of biobanks, three issues regarding data collection and application must be addressed (1) integration and distribution of data governance, (2) efficiency and efficacy of data governance, and (3) sustainability of data governance. Methods: We collaborated with stakeholders to identify priorities and assess infrastructure needs through the continuous evaluation and analysis of projects. We developed data management solutions, catalogs, and data models to optimize and support data collection, distribution, and application. Furthermore, ontologies were used to facilitate data integration from multiple sources, and Minimum Information About BIobank Data Sharing (MIABIS) was defined as accessible to all patients. To enhance data integrity, we conducted retrospective and prospective follow-up studies. Results: We completed infrastructure upgrades to match technical solutions and research demands. An information management software with six primary functional divisions was developed for data governance. We optimized the database structure and changed the biospecimen accumulation model from biospecimen-based to patient-centered and service-oriented. Subsequently, we specified 85 attributes of MIABIS to describe the biobank contents. A dual-pillar approach was adopted to expand the biobank's data in collaboration with other institutions, and MIABIS served as a bridge for both vertical and horizontal networks. From 2003 to 2021, we collected a total of 156,997 patient biospecimens/data from 20 cancer types, matching 53,113 cases from follow-up surveys. In addition, we supplied more than 40,000 biospecimens/data points for above 300 scientific research projects. Conclusions: An appropriate information platform for a biobank is fundamental to data collection, distribution, and application, particularly in the context of data-intensive research. We implemented a standardized scientific data structure to fulfill the research requirements. The sustainable development of a biobank depends on a scientific, standardized, and service-oriented data governance approach, along with the efficient utilization of emerging technologies.

Data Modeling Essentials: Techniques, Best Practices, and Future Trends

Data modeling, in essence, is based on the process of data structuring and organization to effectively handle, analyze, and make informed decisions across organizations. This paper reviews the major concepts and methods significant for modern data architecture in "Data Modeling Essentials: Techniques, Best Practices, and Future Trends.". We start by considering the base techniques, namely Entity-Relationship and Dimensional modeling, and look into their applications in operational and analytical data systems. The paper also covers more advanced topics that include Data Vault modeling for agile data warehousing and NoSQL modeling to accommodate unstructured and semi-structured data, increasingly important in today's big data era. The paper also gives an overview of best practices in developing scalable, flexible data models: stakeholder communication, documentation, and the use of automation tools in accelerating model development and consistency. Data quality, adherence to regulations, and alignment of models with business objectives are among the practices identified as crucial in today's fast-moving data environments. New trends are taking precedence, and data modeling is continuously evolving. We will review the rise of AI- assisted data modeling, graph database structures for highly interconnected data, and model-driven development to accelerate system deployments. This paper examines actionable insights for data professionals, IT teams, and business leaders on how to design robust, agile models that improve the quality of data, enhance decision-making, and foster innovation in an ever-changing data ecosystem. Keywords Data Modeling, Entity-Relationship (ER) Modeling, Dimensional Modeling, Data Quality, Scalable Data Models, Model Driven Development (MDD), Data Management, Structured & Unstructured Data, Future Trends in Data Modeling.

BioDolphin as a comprehensive database of lipid–protein binding interactions

Lipid-protein interactions are crucial for virtually all biological processes in living cells. However, existing structural databases focusing on these interactions are limited to integral membrane proteins. A systematic understanding of diverse lipid-protein interactions also encompassing lipid-anchored, peripheral membrane and soluble lipid binding proteins remains to be elucidated. To address this gap and facilitate the research of universal lipid-protein assemblies, we developed BioDolphin - a curated database with over 127,000 lipid-protein interactions. BioDolphin provides comprehensive annotations, including protein functions, protein families, lipid classifications, lipid-protein binding affinities, membrane association type, and atomic structures. Accessible via a publicly available web server (www.biodolphin.chemistry.gatech.edu), users can efficiently search for lipid-protein interactions using a wide range of options and download datasets of interest. Additionally, BioDolphin features interactive 3D visualization of each lipid-protein complex, facilitating the exploration of structure-function relationships. BioDolphin also includes detailed information on atomic-level intermolecular interactions between lipids and proteins that enable large scale analysis of both paired complexes and larger assemblies. As an open-source resource, BioDolphin enables global analysis of lipid-protein interactions and supports data-driven approaches for developing predictive machine learning algorithms for lipid-protein binding affinity and structures.

Concept and Design of Cutting Tools for Osseodensification in Implant Dentistry

Osseodensification is an innovative surgical instrumentation technique based on additive (non-cutting) drilling using special burs. It is known from the literature, that the osseodensification burs should operate in a clockwise direction to drill holes and in a counterclockwise direction to compact the osteotomy walls. For these purposes, the burs have special design features, like conical contour shape, increased number of helical flutes, and negative rake angle on the peripheral part. However, although other parameters and features of the burs define their overall performance, they are not described sufficiently, and their influence on surgical quality is almost unknown both for clinicians and tool manufacturers. The purpose of the present research is to identify the key design features of burs for osseodensification and their functional relationship with the qualitative indices of the procedure based on an analytical review of research papers and patent documents. It will help to further improve the design of osseodensification burs and thereby enhance the surgical quality and, ultimately, patient satisfaction. Results: The most important design features and parameters of osseodensification burs are identified. Thereon, the structural model of osseodensification bur is first represented as a hypergraph. Based on the analysis of previous research, functional relationships between design parameters of osseodensification burs, osseodensification procedure conditions, and procedure performance data were established and, for the first time, described in the comprehensive form of a hypergraph. Conclusions: This study provides formal models that form the basis of database structure and its control interface, which will be used in the later developed computer-aided design module to create advanced types of burs under consideration. These models will also help to make good experimental designs used in studies aimed at improving the efficiency of the osseodensification procedure.

Exploring the hydrate landscape using data mining on the Cambridge structural database (CSD).

With the continued relevance of drug hydrates in pharmaceutical sciences, a comprehensive understanding of hydrate and anhydrate forms is essential, not only through individual case studies but also from a broader, systematic perspective. The Cambridge Structural Database (CSD) is a well-established database for crystal structures of organic molecules and here, the structural features of pharmaceutically relevant compounds forming hydrates were explored. Drug anhydrate and hydrate subsets were generated and further classified into separate anhydrate and hydrate sets for free drug, cocrystal/solvate, salt, multicomponent cocrystal/solvate, and salt cocrystal/solvate systems. A thorough understanding of these sets was documented at molecular and structural levels. The CSD drug subset contains 24% of entries as hydrates and 76% as anhydrates. Only 6% of anhydrates have corresponding hydrate forms in the CSD drug subset. The formation of hydrates seems to be still less documented in multicomponent drug hydrates, as well as polymorphism of hydrates is less explored for these increasingly complicated systems with a high number of components. The presence of water molecules or additional components does not necessarily lead to a higher degree of crystal packing. Water is involved in 44% of hydrogen bonds (H-bond) in drug hydrate set, where water prefers to act as H-bond donor. H-bonds formed only by water show a relatively high bond strength. This work demonstrates the potential of data science in analyzing pharmaceutically relevant databases to uncover hidden patterns, and more specifically utilizing the CSD for understanding structural aspects and the role of water in H-bond patterns in drug hydrates.