Article
Computer Science, Artificial Intelligence
Balu Bhasuran, Jeyakumar Natarajan
Summary: This study used advanced experimental methods such as next-generation sequencing to identify potential genetic biomarkers and gene variants related to diseases. To extract meaningful information from the scientific literature, sophisticated text mining-based knowledge-driven frameworks were utilized.
KNOWLEDGE AND INFORMATION SYSTEMS
(2023)
Article
Computer Science, Information Systems
Penghua Zhu, Ying Li, Tongyu Li, Huimin Ren, Xiaolei Sun
Summary: Crowdsourced testing is becoming popular in software testing due to its efficiency in utilizing crowdsourcing and cloud platforms. This study focuses on prioritizing test reports in crowdsourced testing by adopting test case prioritization methods. The results demonstrate the effectiveness of these methods with an average APFD of over 0.8.
Article
Health Care Sciences & Services
Svetlana Tarbeeva, Ekaterina Lyamtseva, Andrey Lisitsa, Anna Kozlova, Elena Ponomarenko, Ekaterina Ilgisonis
Summary: By using automatic text-mining, we identified major trends and key genes in the obesity field, reducing a large set of data to 19 genes for potential personalized medicine applications.
JOURNAL OF PERSONALIZED MEDICINE
(2021)
Article
Computer Science, Artificial Intelligence
Sadia Ali, Yaser Hafeez, Shariq Hussain, Shunkun Yang, Muhammad Jamal
Summary: Component-based software development is widely implemented in the current era, aiming to enhance user satisfaction and reduce complexity, but faces challenges such as specification, prioritization, and semantic issues. To address these challenges, a framework using text mining and case-based reasoning techniques is proposed and empirically validated.
Article
Computer Science, Artificial Intelligence
Wei -Chao Lin, Chih-Fong Tsai, Hsuan Chen
Summary: This study utilized text mining techniques and machine learning algorithms for stock market prediction, finding that the combination of CNN with Word2vec and CNN with BERT performed the best. Additionally, the use of different text feature representations and learning models in financial news articles published on different news platforms can have an impact on prediction results.
APPLIED SOFT COMPUTING
(2022)
Article
Immunology
Yoshiro Mori, Nobuyuki Miyatake, Hiromi Suzuki, Yuka Mori, Setsuo Okada, Kiyotaka Tanimoto
Summary: This study compared public perceptions of COVID-19 vaccination and influenza vaccination in Japan by analyzing social media posts on Twitter. It found that on December 15, 2022, there was frequent discussion of death related to the corona vaccine, while severe disease was not mentioned in relation to the influenza vaccine. However, after the recognition of side effects of the corona vaccine, the mentions of death decreased. The comprehensive analysis of social media data revealed distinct variations in public perceptions of the two vaccines in Japan.
Article
Biochemical Research Methods
Senay Kafkas, Marwa Abdelhakim, Mahmut Uludag, Azza Althagafi, Malak Alghamdi, Robert Hoehndorf
Summary: TARVar is an automated tool that ranks variants based on patient symptoms and clinical signs using evidence from literature and genomes. It improves upon existing tools by allowing flexible expression of patient symptoms in free-text format.
BMC BIOINFORMATICS
(2023)
Article
Computer Science, Artificial Intelligence
Christine P. Chai
Summary: This article discusses the importance of text preprocessing and its direct impact on the results of natural language processing applications. The authors explore various common text preprocessing methods and provide examples of special cases that require customized preprocessing. This article serves as a guideline for selecting and fine-tuning text preprocessing methods.
NATURAL LANGUAGE ENGINEERING
(2023)
Article
Computer Science, Information Systems
Xun Li, Lei Liu, Zhiqi Chen, Yuzhou Liu, Huaxiao Liu
Summary: This article introduces the importance of API description and proposes a method (BDBM) that fuses multimodal data to comprehensively describe APIs. The experimental results show that API recommendation based on BDBM outperforms unimodal API information, and BDBM can be used in different API-related tasks.
INFORMATION AND SOFTWARE TECHNOLOGY
(2023)
Article
Biochemistry & Molecular Biology
Haixia Shang, Zhi-Ping Liu
Summary: This paper presents an ensemble method for reliable network-based biomarker discovery, using supervised module detection and module prioritization. The authors successfully identify hepatocellular carcinoma (HCC) network modules as diagnostic biomarkers and validate their effectiveness on gene regulatory networks. The results demonstrate the ability of the method to find effective network biomarkers for cancer diagnosis with fewer false positives.
COMPUTATIONAL AND STRUCTURAL BIOTECHNOLOGY JOURNAL
(2022)
Article
Computer Science, Artificial Intelligence
Mickael Febrissy, Aghiles Salah, Melissa Ailem, Mohamed Nadif
Summary: Non-negative Matrix Factorization (NMF) and its variants are widely used for clustering text documents. However, these methods do not explicitly consider the contextual dependencies between words. To address this issue, researchers draw inspiration from neural word embedding and propose jointly factorizing the document-word and word-word co-occurrence matrices. Empirical results show that this approach significantly improves the clustering performance of NMF on multiple real-world datasets.
Article
Automation & Control Systems
Qin Shi, Yu Zhu, Yatong Liu, Jiongyao Ye, Dawei Yang
Summary: In this paper, a Perceiving Multiple Representations (PerMR) method for scene text image super-resolution is proposed. It combines super-resolution with text recognition and utilizes the feedback of the recognizer to improve super-resolution performance and enhance image quality. Experimental results demonstrate that PerMR can generate more distinguishable images and outperform state-of-the-art methods.
ENGINEERING APPLICATIONS OF ARTIFICIAL INTELLIGENCE
(2023)
Article
Computer Science, Theory & Methods
Mark Abraham Magumba, Peter Nabende
Summary: Twitter and social media can serve as important sources for disease surveillance data, but the messiness of tweets poses challenges for information extraction. Most systems rely on simple keyword matching, leading to potential false positives, and solutions for multilingual scenarios often lack semantic context. The paper experimentally examines different text classification approaches for epidemiological surveillance on the social web and compares the impact of different input representations on performance.
JOURNAL OF BIG DATA
(2021)
Article
Health Care Sciences & Services
Silvia Monaco, Alessia Renzi, Beatrice Galluzzi, Rachele Mariani, Michela Di Trani
Summary: This study explores physiotherapists' perceptions of their relationship with patients and identifies helpful elements for promoting intervention outcomes. Physiotherapists organize their work into three categories: working with the patient, the healing process, and functioning as psychologists. Understanding the emotional and relational processes in physiotherapist practice can lead to integrated interventions that prioritize the individual.
Article
Chemistry, Analytical
Suganya Selvaraj, Eunmi Choi
Summary: Text document clustering involves classifying textual documents into clusters based on content similarity. Swarm intelligence algorithms use simple rules to tackle complex tasks, with PSO and GWO algorithms outperforming K-means in document clustering.
Correction
Computer Science, Interdisciplinary Applications
Edward De Brouwer, Thijs Becker, Yves Moreau, Eva Kubala Havrdova, Maria Trojano, Sara Eichau, Serkan Ozakbas, Marco Onofrj, Pierre Grammond, Jens Kuhle, Ludwig Kappos, Patrizia Sola, Elisabetta Cartechini, Jeannette Lechner-Scott, Raed Alroughani, Oliver Gerlach, Tomas Kalincik, Franco Granella, Francois Grand'Maison, Roberto Bergamaschi, Maria Jose Sa, Bart Van Wijmeersch, Aysun Soysal, Jose Luis Sanchez-Menoyo, Claudio Solaro, Cavit Boz, Gerardo Iuliano, Katherine Buzzard, Eduardo Aguera-Morales, Murat Terzi, Tamara Castillo Trivio, Daniele Spitaleri, Vincent Van Pesch, Vahid Shaygannejad, Fraser Moore, Celia Oreja-Guevara, Davide Maimone, Riadh Gouider, Tunde Csepany, Cristina Ramo-Tello, Liesbet Peeters
COMPUTER METHODS AND PROGRAMS IN BIOMEDICINE
(2022)
Editorial Material
Biochemistry & Molecular Biology
Francesca Forzano, Olga Antonova, Angus Clarke, Guido de Wert, Sabine Hentze, Yalda Jamshidi, Yves Moreau, Markus Perola, Inga Prokopenko, Andrew Read, Alexandre Reymond, Vigdis Stefansdottir, Carla van El, Maurizio Genuardi
Summary: Although polygenic risk score analyses on embryos (PGT-P) are being marketed to parents using in vitro fertilisation as a tool for selecting embryos with lower disease risk, the utility of PRS in this context is limited and lacks clinical research support. Patients need to be informed about the limitations of using PRSs in this way, and a societal debate about the selection of individual traits should take place before further implementation of this technique in this population.
EUROPEAN JOURNAL OF HUMAN GENETICS
(2022)
Article
Biochemistry & Molecular Biology
Daniele Raimondi, Massimiliano Corso, Piero Fariselli, Yves Moreau
Summary: In this paper, a novel Genome Interpretation paradigm called Galiana is proposed, which directly models the genotype-to-phenotype relationship. The model is trained using Whole Genome sequencing data to predict Arabidopsis thaliana phenotypes, particularly related to flowering traits. Galiana achieves better performances and larger phenotype coverage compared to other models, and it is also fully interpretable using Saliency Maps gradient-based approaches. Additionally, 36 novel genes associated with flowering traits are identified.
NUCLEIC ACIDS RESEARCH
(2022)
Article
Multidisciplinary Sciences
Gabriele Orlando, Daniele Raimondi, Ramon Duran-Romana, Yves Moreau, Joost Schymkowitz, Frederic Rousseau
Summary: Structural bioinformatics lacks interfaces connecting with machine learning methods, hindering the application of modern neural network architectures. PyUUL is introduced as a library that translates biological structures into 3D tensors, enabling the application of state-of-the-art deep learning algorithms. The library also supports GPU and sparse calculation, and can be used to address typical bioinformatics problems.
NATURE COMMUNICATIONS
(2022)
Article
Biochemistry & Molecular Biology
Heleen Masset, Jia Ding, Eftychia Dimitriadou, Sophie Debrock, Olga Tsuiko, Katrien Smits, Karen Peeraer, Thierry Voet, Masoud Zamani Esteki, Joris R. Vermeesch
Summary: Single-cell whole-genome haplotyping allows simultaneous detection of haplotypes associated with monogenic diseases, chromosome copy-numbering, and revealed mosaicism in embryos and embryonic stem cells. This sequencing-based method could replace traditional genetic testing methods and has the potential to become more accessible and cost-effective.
NUCLEIC ACIDS RESEARCH
(2022)
Article
Clinical Neurology
Steve Simpson-Yap, Ashkan Pirmani, Tomas Kalincik, Edward De Brouwer, Lotte Geys, Tina Parciak, Anne Helme, Nick Rijke, Jan A. Hillert, Yves Moreau, Gilles Edan, Sifat Sharmin, Tim Spelman, Robert McBurney, Hollie Schmidt, Arnfin B. Bergmann, Stefan Braune, Alexander Stahmann, Rod M. Middleton, Amber Salter, Bruce Bebo, Anneke van der Walt, Helmut Butzkueven, Serkan Ozakbas, Cavit Boz, Rana Karabudak, Raed Alroughani, Juan Rojas, Ingrid A. van der Mei, Guilherme Sciascia do Olival, Melinda Magyari, Ricardo N. Alonso, Richard S. Nicholas, Anibal S. Chertcoff, Ana Zabalza de Torres, Georgina Arrambide, Nupur Nag, Annabel Descamps, Lars Costers, Ruth Dobson, Aleisha Miller, Paulo Rodrigues, Vesna Prckovska, Giancarlo Comi, Liesbet M. Peeters
Summary: This study found that male sex, older age, progressive MS, and higher disability are associated with more severe COVID-19. The use of anti-CD20 medications is also linked to increased severity of COVID-19.
NEUROLOGY-NEUROIMMUNOLOGY & NEUROINFLAMMATION
(2022)
Letter
Biochemistry & Molecular Biology
Francesca Forzano, Olga Antonova, Angus Clarke, Guido de Wert, Sabine Hentze, Yalda Jamshidi, Yves Moreau, Markus Perola, Inga Prokopenko, Andrew Read, Alexandre Reymond, Vigdis Stefansdottir, Carla van El, Maurizio Genuardi, European Soc Human Genetics
EUROPEAN JOURNAL OF HUMAN GENETICS
(2023)
Article
Pharmacology & Pharmacy
Anna Vincze, Gergely Dekany, Richard Bicsak, Andras Formanek, Yves Moreau, Gabor Koplanyi, Gergely Takacs, Gabor Katona, Diana Balogh-Weiser, Adam Arany, Gyorgy T. Balogh
Summary: The permeability of total and polar fractions of bovine heart and liver lipid extracts in the PAMPA model and their relationship with physicochemical descriptors of drug molecules were investigated. The results showed that there were subtle differences between total and polar lipids, while liver lipids had significantly different permeability compared to heart or brain lipid-based models. The study also found correlations between in silico descriptors of drug molecules and permeability values, providing insights into tissue-specific permeability.
Article
Biochemical Research Methods
Eugenio Mazzone, Yves Moreau, Piero Fariselli, Daniele Raimondi
Summary: In this study, we propose a new approach based on data fusion for reliable Drug-Target Interactions (DTIs) prediction. By extending the Matrix Factorization paradigm to the nonlinear inference over Entity-Relation graphs using the NXTfusion library, our models outperform most existing methods and have the flexibility to predict both binary DTIs and regression of real-valued drug-target affinity. Our findings suggest that DTI methods should be validated in settings that mimic real-life situations where predictions for previously unseen drugs, proteins, and drug-protein pairs are needed. Integration of heterogeneous information with our Entity-Relation data fusion approach is most beneficial in such contexts.
Article
Medical Informatics
Ashkan Pirmani, Edward De Brouwer, Lotte Geys, Tina Parciak, Yves Moreau, Liesbet M. Peeters
Summary: This study presents a comprehensive data analysis pipeline driven by multiple stakeholders, which accommodates three prevalent data-sharing streams and has been successfully implemented in the global data sharing initiatives for multiple sclerosis and COVID-19. The pipeline facilitates data gathering from various sources and integrates them into a unified dataset for subsequent statistical analysis and secure data examination.
JMIR MEDICAL INFORMATICS
(2023)
Article
Mathematics, Applied
Susan Ghaderi, Masoud Ahookhosh, Adam Arany, Alexander Skupin, Panagiotis Patrinos, Yves Moreau
Summary: This paper proposes a gradient-based Markov Chain Monte Carlo (MCMC) method for sampling from the posterior distribution of problems with nonsmooth potential functions. By using smoothing techniques, the original potential function is approximated by a smooth function with the same critical points, leading to a smoothing ULA method called SULA. Non-asymptotic convergence results of SULA are established under mild assumptions on the original potential function. Numerical results demonstrate the promising performance of SULA on both synthetic and real chemoinformatics data.
APPLIED MATHEMATICS AND COMPUTATION
(2024)
Meeting Abstract
Clinical Neurology
S. Simpson-Yap, E. De Brouwer, T. Kalincik, N. Rijke, J. Hillert, C. Walton, G. Edan, Y. Moreau, T. Spelman, L. Peeters
MULTIPLE SCLEROSIS JOURNAL
(2022)
Meeting Abstract
Clinical Neurology
S. Simpson-Yap, E. De Brouwer, T. Kalincik, N. Rijke, J. Hillert, C. Walton, G. Edan, Y. Moreau, T. Spelman, L. Peeters
MULTIPLE SCLEROSIS JOURNAL
(2022)
Article
Biochemistry & Molecular Biology
Daniele Raimondi, Francesco Codice, Gabriele Orlando, Joost Schymkowitz, Frederic Rousseau, Yves Moreau
Summary: This article presents HPMPdb, a database containing detailed descriptions of human Single Amino acid Variants (SAVs) and their effects on protein molecular phenotypes. The database allows researchers to go beyond the existing Deleterious/Neutral prediction paradigm and build molecular phenotype predictors. Necessary means for training and testing models on the database are provided.
CURRENT RESEARCH IN STRUCTURAL BIOLOGY
(2022)
Proceedings Paper
Computer Science, Artificial Intelligence
Jaak Simm, Adam Arany, Edward De Brouwer, Yves Moreau
Summary: This paper introduces a route-based multi-attention mechanism that incorporates features from routes between node pairs, aiming at addressing the information bottleneck issue in deep learning from molecular graphs. The proposed method, called Graph Informer, is able to attend to nodes several steps away, and it outperforms existing approaches in two prediction tasks. Furthermore, a variant method called injective Graph Informer is developed and proven to be more powerful than the Weisfeiler-Lehman test for graph isomorphism.
MACHINE LEARNING, OPTIMIZATION, AND DATA SCIENCE (LOD 2021), PT II
(2022)