Article
Computer Science, Software Engineering
Bharat Kale, Austin Clyde, Maoyuan Sun, Arvind Ramanathan, Rick Stevens, Michael E. E. Papka
Summary: Exploratory analysis of chemical space is crucial but impractical to be done manually. ChemoGraph is a novel visual analytics technique that formalizes chemical space as a hypergraph and utilizes machine learning models to compute related compounds, enabling the enlargement of the known space. It also provides interactive features for users to view, compare, and organize related chemicals.
COMPUTER GRAPHICS FORUM
(2023)
Article
Biochemical Research Methods
Tao Zeng, Bernard Andes Hess, Fan Zhang, Ruibo Wu
Summary: A bio-inspired strategy named TeroGen is developed to mimic the two key biosynthetic stages of terpenoid natural products using physically based simulations and deep learning models. It can predict and estimate the synthetic accessibility and chemical interpretation of tens of thousands of sesterterpenoids, thereby expanding the chemical space of terpenoids.
BRIEFINGS IN BIOINFORMATICS
(2022)
Article
Chemistry, Multidisciplinary
Chong Lu, Shien Liu, Weihua Shi, Jun Yu, Zhou Zhou, Xiaoxiao Zhang, Xiaoli Lu, Faji Cai, Ning Xia, Yikai Wang
Summary: SECSE is a novel gene design platform that integrates artificial intelligence and deep learning to efficiently search and generate small molecules with potential drug activity.
JOURNAL OF CHEMINFORMATICS
(2022)
Article
Computer Science, Software Engineering
Soumaya Rebai, Vahid Alizadeh, Marouane Kessentini, Houcem Fehri, Rick Kazman
Summary: This paper proposes an interactive approach that allows developers to pinpoint their preferences in both the objective and decision spaces, resulting in more efficient resolution of quality issues. By using multi-objective search and clustering algorithms, developers are able to examine a smaller number of solutions and provide feedback, which is then used to generate constraints and optimize the refactoring process.
IEEE TRANSACTIONS ON SOFTWARE ENGINEERING
(2022)
Article
Computer Science, Software Engineering
Masaaki Miki, Toby Mitchell
Summary: Achieving a pure-compression stress state is important in form-finding of shell structures. However, this assumption restricts the geometry of the structure's plan, and allowing both tension and compression becomes essential. The compatibility of boundary conditions is a challenge when performing tension-compression mixed form-finding.
ACM TRANSACTIONS ON GRAPHICS
(2022)
Article
Biochemistry & Molecular Biology
Philip Z. Johnson, Anne E. Simon
Summary: RNAcanvas is a tool for drawing nucleic acid structures, with automatic arrangement and interactive editing features. It supports customizable elements and robust performance optimizations for large structures. It also enables real-time highlighting of complementary sequences and motif search, aiding in analyzing structures and making comparisons. The drawings can be exported in both SVG and PowerPoint formats for publication quality figures.
NUCLEIC ACIDS RESEARCH
(2023)
Review
Engineering, Biomedical
Mae Jemison, Ronke Olabisi
Summary: This work reviews the major human and environmental challenges in human spaceflight, and how biomaterials could help address some of these challenges. It highlights the significance of biomaterials in space programs and the potential for advancing biomaterial technology on Earth. Additionally, it explores the fabrication of novel biomaterials in space environments and how these technologies could have implications for Earth.
ACTA BIOMATERIALIA
(2021)
Article
Computer Science, Cybernetics
Katerina Vitsaxaki, Stavroula Ntoa, George Margetis, Nicolas Spyratos
Summary: This paper proposes and implements a tool for accessing, analyzing, and visually exploring large relational databases and data warehouses. The tool represents the database as a graph and allows users to define analytic queries through a user interface, which are then translated into SQL group-by queries. The tool also provides visualization of the analysis results and allows users to interact with the visualizations to explore the database content.
INTERNATIONAL JOURNAL OF HUMAN-COMPUTER INTERACTION
(2023)
Article
Computer Science, Software Engineering
Glen Berseth, Brandon Haworth, Muhammad Usman, Davide Schaumann, Mahyar Khayatkhoei, Mubbasir Kapadia, Petros Faloutsos
Summary: IDOME is an interactive system for computer-aided design optimization, striking a balance between automation and control to assist architects in exploring and optimizing space layouts efficiently. The system provides alternative building layouts that meet user-defined constraints and optimization criteria, allowing users to iterate the design exploration process.
IEEE TRANSACTIONS ON VISUALIZATION AND COMPUTER GRAPHICS
(2021)
Review
Biology
Siddhant Sharma, Aayush Arya, Romulo Cruz, Henderson Cleaves
Summary: The study of prebiotic chemistry involves complex systems of chemical reactions that may have led to the origin of life on Earth. Experimentalists find it difficult to study these systems in laboratory simulations, but computational chemistry offers efficient ways to analyze and identify self-replicating features central to the origin of life models. By modeling prebiotic chemical reaction networks, researchers can better understand the emergence of life.
Article
Materials Science, Multidisciplinary
Axel van de Walle, Hantong Chen, Helena Liu, Chiraag Nataraj, Sayan Samanta, Siya Zhu, Raymundo Arroyave
Summary: High-dimensional thermodynamic phase stability databases are increasingly common, addressing the need for intuitive understanding of phase relationships in materials design through algorithms that enable interactive exploration in high-dimensional spaces.
Article
Mathematics, Applied
Stefan Buschmann, Peter Hoffmann, Ankit Agarwal, Norbert Marwan, Thomas Nocke
Summary: This paper introduces Geo-Temporal eXplorer (GTX), a GPU-based tool for visual analytics of large geo-referenced complex networks in the climate research domain. It discusses solutions for interactive visual analysis of various types of large complex networks, including time-dependent, multi-scale, and multi-layered ensemble networks. The GTX tool supports heterogeneous tasks and provides interactive GPU-based solutions for real-time processing, analysis, and visualization of large network data.
Article
Computer Science, Software Engineering
Yifang Wang, Hongye Liang, Xinhuan Shu, Jiachen Wang, Ke Xu, Zikun Deng, Cameron Campbell, Bijia Chen, Yingcai Wu, Huamin Qu
Summary: The availability of quantitative historical datasets has opened up new research opportunities for various social science disciplines. However, existing statistical approaches are not suitable for analyzing career mobility in historical datasets due to their fine-grained attributes and long time span. This article proposes an interactive visual analytics system, CareerLens, to assist experts in exploring and understanding historical career data.
IEEE TRANSACTIONS ON VISUALIZATION AND COMPUTER GRAPHICS
(2022)
Article
Computer Science, Software Engineering
Laura Garrison, Juliane Mueller, Stefanie Schreiber, Steffen Oeltze-Jafra, Helwig Hauser, Stefan Bruckner
Summary: DimLift is a novel visual analysis method for creating and interacting with dimensional bundles. Dimensional bundles are expressive groups of dimensions that contribute similarly to the variance of a dataset. Through interactive exploration and reconstruction methods, users can lift interesting and subtle relationships to the surface.
IEEE TRANSACTIONS ON VISUALIZATION AND COMPUTER GRAPHICS
(2021)
Article
Engineering, Biomedical
Benedikt Mayer, Monique Meuschke, Jimmy Chen, Beat P. Muller-Stich, Martin Wagner, Bernhard Preim, Sandy Engelhardt
Summary: This article introduces a web-based application design for analyzing surgical procedure recordings and gaining insights into surgical workflows and datasets. The evaluation of the application showed its complexity as an expert tool, but it allowed users to correctly solve various analysis tasks and come up with novel hypotheses regarding the data.
INTERNATIONAL JOURNAL OF COMPUTER ASSISTED RADIOLOGY AND SURGERY
(2023)
Article
Chemistry, Medicinal
Jocelyn Sunseri, David R. Koes
JOURNAL OF CHEMICAL INFORMATION AND MODELING
(2020)
Meeting Abstract
Biochemistry & Molecular Biology
Faiha Khan, Di Zhang, David Koes
Article
Biochemistry & Molecular Biology
David Gau, Lucile Vignaud, Abigail Allen, Zhijian Guo, Jose Sahel, David Boone, David Koes, Xavier Guillonneau, Partha Roy
JOURNAL OF BIOLOGICAL CHEMISTRY
(2020)
Article
Chemistry, Physical
Dakota L. Folmsbee, David R. Koes, Geoffrey R. Hutchison
Summary: The study evaluates the accuracy of several mainstream machine learning methods and finds that some of them offer qualitative and quantitative accuracy on small molecules, such as correct minima and mean absolute percent error. ANI-2x, FCHL, and a new libmolgrid-based convolutional neural net, the Colorful CNN, show good performance.
JOURNAL OF PHYSICAL CHEMISTRY A
(2021)
Correction
Chemistry, Medicinal
Paul G. Francoeur, David R. Koes
JOURNAL OF CHEMICAL INFORMATION AND MODELING
(2021)
Article
Chemistry, Medicinal
Paul G. Francoeur, David R. Koes
Summary: SolTranNet is a machine learning model for predicting aqueous solubility based on a molecule's SMILES representation. Contrary to previous assumptions, the study found that larger models do not necessarily perform better, with SolTranNet outperforming linear ML approaches. The model shows promising results and competitive performance when used to filter out insoluble compounds.
JOURNAL OF CHEMICAL INFORMATION AND MODELING
(2021)
Article
Biochemistry & Molecular Biology
Jonathan Edward King, David Ryan Koes
Summary: This article introduces a new dataset, SidechainNet, which extends ProteinNet and includes angle and atomic coordinate information for protein structures. By incorporating sidechain information, SidechainNet can describe all heavy atoms, and users can extend it to include new protein structures.
PROTEINS-STRUCTURE FUNCTION AND BIOINFORMATICS
(2021)
Article
Chemistry, Multidisciplinary
Andrew T. McNutt, Paul Francoeur, Rishal Aggarwal, Tomohide Masuda, Rocco Meli, Matthew Ragoza, Jocelyn Sunseri, David Ryan Koes
Summary: Molecular docking software Gnina 1.0, utilizing convolutional neural networks as scoring functions, outperforms AutoDock Vina in redocking and cross-docking tasks when binding pockets are explicitly defined. The ensemble of CNNs shows good generalization to unseen proteins and ligands, producing scores that correlate well with known binding poses. The 1.0 version of GNINA is available under an open source license for use as a molecular docking tool.
JOURNAL OF CHEMINFORMATICS
(2021)
Article
Ophthalmology
David Gau, Lucile Vignaud, Paul Francoeur, David Koes, Xavier Guillonneau, Partha Roy
Summary: Aberrant angiogenesis is at the core of many ocular pathologies, and this study focuses on a novel small molecule compound that has antiangiogenic activity by inhibiting profilin1-actin interaction. The compound was found to inhibit migration, proliferation, and angiogenic activity of microvascular endothelial cells in vitro, as well as choroidal neovascularization ex vivo. Preliminary structure-activity relationship study showed potential for improving biological activity through structural modifications.
EXPERIMENTAL EYE RESEARCH
(2021)
Article
Biochemistry & Molecular Biology
Jocelyn Sunseri, David Ryan Koes
Summary: Virtual screening is important in drug discovery, but there are inherent tradeoffs between accuracy and speed in the algorithms used. The Gnina molecular docking software outperforms conventional empirical scoring in virtual screening, but bias issues still exist.
Article
Chemistry, Multidisciplinary
Matthew Ragoza, Tomohide Masuda, David Ryan Koes
Summary: This study presents a deep learning system for generating 3D molecular structures conditioned on a receptor binding site, using an atomic density grid representation to train a conditional variational autoencoder. The properties of the generated molecules are evaluated, showing significant changes under different conditions such as mutated receptors. Sampling and interpolation techniques are used to explore the latent space learned by the generative model, allowing for end-to-end prediction of stable bioactive molecules from protein structures with deep learning.
Article
Chemistry, Medicinal
Andrew T. McNutt, David Ryan Koes
Summary: In this study, a Siamese convolutional neural network (CNN) is proposed for the prediction of relative binding free energy (RBFE) between two bound ligands. The network shows improved performance in RBFE prediction compared to a standard CNN, and its predictive performance varies among different protein families. Additionally, the RBFE prediction performance can be enhanced through few-shot learning during model training.
JOURNAL OF CHEMICAL INFORMATION AND MODELING
(2022)
Article
Chemistry, Multidisciplinary
Harrison Green, David R. Koes, Jacob D. Durrant
Summary: Machine learning has been increasingly applied in the field of computer-aided drug discovery, showing notable advances in binding-affinity prediction, virtual screening, and QSAR. A deep convolutional neural network was used to predict appropriate fragments based on the structure of a receptor/ligand complex, with an efficiency of about 58% in selecting correct fragments from known ligands. The trained DeepFrag model and its associated software have been released under the Apache License, Version 2.0.
Article
Medicine, Research & Experimental
Zoltan N. Oltvai, Susan E. Harley, David Koes, Stephen Michel, Erica D. Warlick, Andrew C. Nelson, Sophia Yohe, Pawel Mroz
Summary: This study describes the development of drug resistance in an AML patient treated with the IDH1 inhibitor ivosidenib. Monitoring for acquired drug resistance through full-exon NGS and structural modeling can help in choosing follow-up therapy for IDH1 inhibitor-treated AML patients.
COLD SPRING HARBOR MOLECULAR CASE STUDIES
(2021)