Article
Chemistry, Applied
Ganesan Raman
Summary: This study uses machine learning to predict extra-large pore structures in zeolites based on a large collection of synthetic records. XGBoost algorithm shows good accuracy in predicting these structures and helps in understanding the chemistry of extra-large pore zeolites. The trained XGBoost model is deployed on the Heroku cloud platform and achieves over 85% accuracy in matching experimental results.
MICROPOROUS AND MESOPOROUS MATERIALS
(2023)
Article
Computer Science, Software Engineering
Zhe Yu, Fahmid Morshed Fahid, Huy Tu, Tim Menzies
Summary: Keeping track of and managing Self-Admitted Technical Debts (SATDs) are crucial for maintaining a healthy software project. The proposed two-step framework Jitterbug offers a more efficient approach to identifying SATDs, reducing human effort.
IEEE TRANSACTIONS ON SOFTWARE ENGINEERING
(2022)
Article
Chemistry, Multidisciplinary
Eli Nimy, Moeketsi Mosia, Colin Chibaya
Summary: The utilization of learning analytics to identify at-risk students for early intervention has shown promising results. This study incorporated probabilistic machine learning to include domain knowledge and quantify uncertainty in model parameters and predictions. The study developed a five-stage, probabilistic logistic regression model to identify at-risk students at different stages throughout the academic calendar.
APPLIED SCIENCES-BASEL
(2023)
Article
Chemistry, Analytical
Jared S. Stine, Nicolas Aziere, Bryan J. Harper, Stacey L. Harper
Summary: As plastic production increases, plastic waste accumulates and degrades into smaller plastic particles known as nanoplastics. These nanoplastics are expected to exist in much larger quantities than microplastics. However, detecting nanoplastics in the environment is challenging due to their small size and low mass. This research aims to adapt existing tools to overcome the analytical challenges of identifying nanoplastics. The study proposes using scanning electron microscopy to collect spatiotemporal deformation data and differentiate nanoplastics based on polymer type.
Article
Biochemical Research Methods
Jing Xu, Fuyi Li, Chen Li, Xudong Guo, Cornelia Landersdorfer, Hsin-Hui Shen, Anton Y. Peleg, Jian Li, Seiya Imoto, Jianhua Yao, Tatsuya Akutsu, Jiangning Song
Summary: Antimicrobial peptides (AMPs) are short peptides with various functional activities against target organisms and have the potential to be alternatives to antibiotics in the face of increasing antibiotic resistance. Existing computational approaches for identifying AMPs lack the ability to predict functional activities comprehensively. In this study, we developed a deep learning-based framework, iAMPCN, which significantly improved the prediction performance of AMPs and their functional activities. The model outperformed state-of-the-art approaches and can be used as a valuable tool for identifying potential AMPs with specific functional activities.
BRIEFINGS IN BIOINFORMATICS
(2023)
Article
Medical Informatics
Emily R. Pfaff, Andrew T. Girvin, Tellen D. Bennett, Abhishek Bhatia, Ian M. Brooks, Rachel R. Deer, Jonathan P. Dekermanjian, Sarah Elizabeth Jolley, Michael G. Kahn, Kristin Kostka, Julie A. McMurry, Richard Moffitt, Anita Walden, Christopher G. Chute, Melissa A. Haendel
Summary: Long COVID has had a severe impact on patient and societal recovery from the COVID-19 pandemic. This study developed machine learning models to accurately identify potential long COVID patients using electronic health records. Important features in identifying long COVID included healthcare utilization rate, patient age, dyspnea, and other diagnosis and medication information.
LANCET DIGITAL HEALTH
(2022)
Article
Computer Science, Information Systems
Yufang Huang, Yifan Liu, Peter A. D. Steel, Kelly M. Axsom, John R. Lee, Sri Lekha Tummalapalli, Fei Wang, Jyotishman Pathak, Lakshminarayanan Subramanian, Yiye Zhang
Summary: DICE is a self-supervised learning framework that can identify clinically similar and risk-stratified subgroups with superior performance metrics and predictive power. Clinical evaluation shows that DICE-generated subgroups have predictive value for outcome prediction.
JOURNAL OF THE AMERICAN MEDICAL INFORMATICS ASSOCIATION
(2021)
Article
Environmental Sciences
Sabine Oskar, Mary S. Wolff, Susan L. Teitelbaum, Jeanette A. Stingone
Summary: This study utilized a two-stage machine learning approach to analyze the relationship between exposure to multiple EDCs during childhood and the timing of menarche, identifying various exposure biomarker combinations associated with earlier menarche.
ENVIRONMENTAL RESEARCH
(2021)
Article
Biochemical Research Methods
Xiaoyu Wang, Fuyi Li, Jing Xu, Jia Rong, Geoffrey Webb, Zongyuan Ge, Jian Li, Jiangning Song
Summary: Protein secretion is crucial for intercellular communication, and the non-classical secretion pathway is an important mechanism. This study proposes a new deep learning-based framework that improves the prediction of non-classical secreted proteins.
BRIEFINGS IN BIOINFORMATICS
(2022)
Article
Biochemical Research Methods
Ewan Carr, Mathieu Carriere, Bertrand Michel, Frederic Chazal, Raquel Iniesta
Summary: This paper introduces a clustering pipeline based on Mapper that can identify and summarize clusters using statistically significant topological features.
BMC BIOINFORMATICS
(2021)
Article
Engineering, Multidisciplinary
El-Sayed Atlam, Ashraf Ewis, M. M. Abd El-Raouf, Osama Ghoneim, Ibrahim Gad
Summary: COVID-19 pandemic has had a significant impact on education systems and the psychological health of university students worldwide. This study used statistics and machine learning approaches to examine the effects of online education during COVID-19 on students' psychological health and academic performance, revealing correlations and positive relationships.
ALEXANDRIA ENGINEERING JOURNAL
(2022)
Article
Psychiatry
Patrick Gagnon-Sanschagrin, Jeff Schein, Annette Urganus, Elizabeth Serra, Yawen Liang, Primrose Musingarimi, Martin Cloutier, Annie Guerin, Lori L. Davis
Summary: This study used machine learning to identify potentially undiagnosed PTSD patients in the US commercial population, providing insights into patient characteristics, symptoms, treatments, and healthcare resource utilization.
Article
Environmental Sciences
Mengyuan Li, Zhilan Zhang, Wenxiu Cao, Yijing Liu, Beibei Du, Canping Chen, Qian Liu, Md Nazim Uddin, Shanmei Jiang, Cai Chen, Yue Zhang, Xiaosheng Wang
Summary: The study found that factors such as economic inequality, blood type, disease prevalence, and nutrient intake are closely related to the risk and transmission of COVID-19. High temperature has a more significant effect on reducing COVID-19 transmission.
SCIENCE OF THE TOTAL ENVIRONMENT
(2021)
Article
Genetics & Heredity
Nicholas Donnelly, Adam Cunningham, Sergio Marco Salas, Matthew Bracher-Smith, Samuel Chawner, Jan Stochl, Tamsin Ford, F. Lucy Raymond, Valentina Escott-Price, Marianne B. M. van den Bree
Summary: This study developed a screening tool using machine learning techniques to identify individuals with genomic conditions associated with neurodevelopmental disorders (ND-GCs) who would benefit from additional support. The study identified a set of variables that accurately differentiate individuals with ND-GCs from controls and identified 5 dimensions related to behavior, anxiety, communication, and motor development. This research provides an important foundation for screening and assessing young people with ND-GCs.
Article
Chemistry, Physical
Nozar Moradi, Mohammad Hadi Tavana, Mohammad Reza Habibi, Moslem Amiri, Mohammad Javad Moradi, Visar Farhangi
Summary: Supplementary cementitious materials (SCMs) are widely used in the concrete industry due to their advantages. Different SCMs with various characteristics can enhance the mechanical properties of concrete. By using combinations of two or more SCMs, the multiple demands of concrete properties can be addressed. This study aims to develop a robust and time-saving method based on Machine Learning (ML) to predict the compressive strength of concrete containing binary SCMs at various ages.
Article
Chemistry, Physical
Sibel Ozkaya, Estela Blaisten-Barojas
Article
Multidisciplinary Sciences
Douglas M. Reitz, Estela Blaisten-Barojas
SCIENTIFIC REPORTS
(2019)
Article
Biochemistry & Molecular Biology
Gideon K. Gogovi, Fahad Almsned, Nicole Bracci, Kylene Kehn-Hall, Amarda Shehu, Estela Blaisten-Barojas
Article
Chemistry, Physical
James Andrews, Estela Blaisten-Barojas
JOURNAL OF PHYSICAL CHEMISTRY B
(2019)
Article
Nanoscience & Nanotechnology
Scott D. Hopkins, Gideon K. Gogovi, Eric Weisel, Robert A. Handler, Estela Blaisten-Barojas
Article
Polymer Science
James Andrews, Robert A. Handler, Estela Blaisten-Barojas
Article
Chemistry, Medicinal
Srilatha Sakamuru, Jinghua Zhao, Menghang Xia, Huixiao Hong, Anton Simeonov, Iosif Vaisman, Ruili Huang
Summary: Novel computational models were developed to predict the activity of opioid receptors based on chemical structures, successfully identifying new active compounds. Experimental validation showed that the best performing model, using the random forest classifier, achieved hit rates ranging from 2.3% to 15.8%, enriching hit rates by >= 2-fold compared to the original assay.
JOURNAL OF CHEMICAL INFORMATION AND MODELING
(2021)
Article
Chemistry, Physical
James Andrews, Estela Blaisten-Barojas
Summary: This study uses atomistic simulations to investigate the interface interactions between solid PLGA and water, ethyl acetate, and a mixture of both, as well as exploring the formation of macromolecular assemblies at the PLGA-solvent interface with the addition of DSPE-PEG. The adhesion of nanopatches to the surface is driven by dispersive forces, while keeping the solvent around the new formations is dominated by electrostatic forces. The predicted mechanism of PEG-lipid nanopatch formation could be generally applicable for tailoring asymmetric PLGA nanoparticles for specific drug delivery conditions.
JOURNAL OF PHYSICAL CHEMISTRY B
(2022)
Article
Chemistry, Physical
Daniel Sponseller, Estela Blaisten-Barojas
Summary: Extensive all-atom molecular dynamics studies were conducted on polyethylene glycol (PEG(2000)) in solvated and polymer bulk condensed phases, revealing structural changes under different conditions, with predicted properties aligning well with experiments, laying a foundation for the preparation of novel composite materials in the future.
JOURNAL OF PHYSICAL CHEMISTRY B
(2021)
Article
Physics, Condensed Matter
Yoseph Abere, Greg Helmick, Estela Blaisten-Barojas
Summary: A novel model potential is developed for simulating oxidised oligopyrroles in condensed phases. The optimized force field parameters show excellent agreement with experimental results.
JOURNAL OF PHYSICS-CONDENSED MATTER
(2022)
Article
Biochemistry & Molecular Biology
Kianoush Jeiran, Scott M. Gordon, Denis O. Sviridov, Angel M. Aponte, Amanda Haymond, Grzegorz Piszczek, Diego Lucero, Edward B. Neufeld, Iosif I. Vaisman, Lance Liotta, Ancha Baranova, Alan T. Remaley
Summary: This study developed a novel method to divide apoB-100 into subunits and domains, and validated the models using mass spectrometry cross-linking and known disulfide bond positions. The continuous structure of apoB-100 was generated, and the dynamics during particle size transitions were examined. Additionally, the proposed model of receptor ligand binding provides new insights into the functions of apoB-100.
INTERNATIONAL JOURNAL OF MOLECULAR SCIENCES
(2022)
Article
Chemistry, Multidisciplinary
James Andrews, Olga Gkountouna, Estela Blaisten-Barojas
Summary: This paper explores the ability of three recurrent neural network architectures to predict the energetics of a liquid solution. The results show that these architectures can accurately reproduce the time series, but are limited in their ability to forecast energy in the short or long term. The authors propose a computational protocol that improves energy forecasting accuracy by utilizing time patterns. Although the distribution of points in the energy forecast band is not optimal, the proposed protocol provides useful estimates.
Article
Multidisciplinary Sciences
Majid Masso, Arnav Bansal, Preethi Prem, Akhil Gajjala, Iosif I. Vaisman
Article
Plant Sciences
Joachim Fisahn, Peter Barlow, Gerhard Dorda
Article
Computer Science, Interdisciplinary Applications
Robert A. Handler, Estela Blaisten-Barojas, Phillip M. Ligrani, Pei Dong, Mikell Paige
ENGINEERING REPORTS
(2020)