Article
Biochemical Research Methods
Arika Fukushima, Masahiro Sugimoto, Satoru Hiwa, Tomoyuki Hiroyasu
Summary: This study proposed a method that integrates time-course gene expression profiles for prediction and gene selection, accurately predicting patient response to therapy and successfully selecting genes associated with disease pathology.
BMC BIOINFORMATICS
(2021)
Article
Biochemical Research Methods
Alisa Pavel, Giusy del Giudice, Michele Fratello, Leo Ghemtio, Antonio Di Lieto, Jari Yli-Kauhaluoma, Henri Xhaard, Antonio Federico, Angela Serra, Dario Greco
Summary: In this study, a network mapping approach (KNeMAP) was proposed for comparing transcriptomic profiles. By grouping genes into similarity groups based on multiple levels of prior information, KNeMAP provided a higher-level view and showed higher accuracy in identifying compounds with similar molecular responses compared to traditional methods.
Article
Plant Sciences
Alan Flores-Diaz, Christian Escoto-Sandoval, Felipe Cervantes-Hernandez, Jose J. Ordaz-Ortiz, Corina Hayano-Kanashiro, Humberto Reyes-Valdes, Ana Garces-Claver, Neftali Ochoa-Alejo, Octavio Martinez
Summary: Gene co-expression networks are powerful tools for understanding functional interactions between genes. We present an algorithm to construct gene functional networks for genes annotated in a given biological process or other aspects of interest. The algorithm is based on the correlation of time expression profiles and ensures robustness by repeatedly finding gene expression relations in independent genotypes. Additionally, we propose an algorithm to identify transcription factor candidates for regulating hub genes within the network.
Article
Biochemical Research Methods
Ningyi Zhang, Tianyi Zang
Summary: ImpAESim focuses on extracting a low-dimensional vector representation of features based on ncRNA regulation and gene-gene interaction network. Our method can significantly reduce the calculation bias resulted from the sparse disease associations derived from semantic associations.
BMC BIOINFORMATICS
(2022)
Article
Computer Science, Information Systems
Jin Zhu, Dayu Cheng, Weiwei Zhang, Ci Song, Jie Chen, Tao Pei
Summary: This paper proposes an accurate and reasonable indoor trajectory similarity measure, ISTSM, which considers the features of indoor trajectories and indoor semantic information simultaneously. Experimental results demonstrate that ISTSM is more accurate and reasonable compared to other popular trajectory similarity measures.
ISPRS INTERNATIONAL JOURNAL OF GEO-INFORMATION
(2021)
Article
Multidisciplinary Sciences
Pietro E. Cippa, Federica Cugnata, Paolo Ferrari, Chiara Brombin, Lorenzo Ruinelli, Giorgia Bianchi, Nicola Beria, Lukas Schulz, Enos Bernasconi, Paolo Merlani, Alessandro Ceschi, Clelia Di Serio
Summary: This study, based on comprehensive monitoring of 576 COVID-19 patients, utilized different statistical approaches to gain insights on risk factor identification and dependency structure among clinical and demographic characteristics. The results indicated a significant role of RAASi in reducing the risk of death in COVID-19 patients, suggesting the need for further investigation in prospective randomized controlled trials.
PROCEEDINGS OF THE NATIONAL ACADEMY OF SCIENCES OF THE UNITED STATES OF AMERICA
(2021)
Article
Geochemistry & Geophysics
Yuan Wu, Xiaolei Li, Qingjie Gong, Xuan Wu, Ning Yao, Cheng Peng, Yuede Chao, Xuyang Wang, Xiulang Pu
Summary: Through analyzing samples from two weathering profiles in Dunhua, Jilin Province, the geochemical lithogene LG01 demonstrates good heredity and inheritance, indicating stability during weathering. Furthermore, the validity of the composition classification method based on LG01's acidic similarity was verified.
APPLIED GEOCHEMISTRY
(2021)
Article
Business
Daniel S. Hain, Roman Jurowetzki, Tobias Buchmann, Patrick Wolf
Summary: This paper presents an efficiently scalable approach to measuring technological similarity between patents by combining embedding techniques and nearest-neighbor approximation. It demonstrates the usefulness of this methodology in measuring knowledge flows, mapping technological change, and creating patent quality indicators using the case of electric vehicle technologies. The paper contributes to the growing literature on text-based indicators for patent analysis.
TECHNOLOGICAL FORECASTING AND SOCIAL CHANGE
(2022)
Article
Biochemical Research Methods
Zhixia Teng, Linyue Shi, Haihao Yu, Chengyan Wu, Zhen Tian
Summary: Long non-coding RNAs (lncRNAs) are essential non-coding RNAs with a length greater than 200 nts. Computational-based approaches have been employed to measure the functional similarity of lncRNAs due to the time-consuming and labor-intensive nature of wet-experiments. In this study, a novel approach called MFSLNC was proposed to comprehensively measure the functional similarity of lncRNAs based on variable k-mer profiles of nucleotide sequences.
Article
Computer Science, Information Systems
Muhammad Jawad Hussain, Heming Bai, Yuncheng Jiang
Summary: This paper presents a model called WBLM that improves the link-based vector representations of concepts by exploring the Wikipedia link structure as a semantic graph and assigning weights to connected links. Experimental results show that the proposed WBLM model significantly improves the SS and SR computation accuracy of the WOLM and WTLM models.
INFORMATION PROCESSING & MANAGEMENT
(2023)
Article
Remote Sensing
Neda Kaffash Charandabi, Abolghasem Sadeghi-Niaraki, Soo-Mi Choi, Tamer Abuhmed
Summary: This study compares the COVID-19 situation in different cities to identify similar cities and analyze the impact of proximity factors, weather, and demographics on their similarity. The results demonstrate the pairs of similar cities and the range of differences among various criteria, providing important guidance for choosing future outbreak control strategies.
GEO-SPATIAL INFORMATION SCIENCE
(2023)
Article
Biochemical Research Methods
Yunpei Xu, Hong-Dong Li, Yi Pan, Feng Luo, Fang-Xiang Wu, Jianxin Wang
Summary: Single-cell RNA sequencing (scRNA-seq) technology provides gene expression profiles at single-cell resolution, allowing researchers to explore cell population heterogeneity and genetic variability. One current research direction is accurately identifying different cell types through unsupervised clustering methods.
IEEE-ACM TRANSACTIONS ON COMPUTATIONAL BIOLOGY AND BIOINFORMATICS
(2021)
Article
Oncology
Mateusz Sikora, Katarzyna Krajewska, Klaudia Marcinkowska, Anna Raciborska, Rafal Jakub Wiglusz, Agnieszka Smieszek
Summary: Osteosarcoma is a malignant bone tumor that mainly affects children and elderly people. The primary treatment methods for osteosarcoma are still underdeveloped, leading to poor survival rates. This study aimed to establish the expression profile of molecular markers related to osteosarcoma survival, progression, and metastasis. The findings may facilitate the selection of reliable cellular models for pre-clinical investigations on osteosarcoma therapy development.
Article
Plant Sciences
Xiangshang Song, Yaping Kou, Mingao Duan, Bo Feng, Xiaoyun Yu, Ruidong Jia, Xin Zhao, Hong Ge, Shuhua Yang
Summary: In this study, 25 RcSWEET genes were identified in Rosa chinensis 'Old Blush' and characterized based on their phylogenetic relationships, gene structures, and regulatory roles during stress and hormone responses. Tissue-specific and cold-response expression profiles showed the diverse roles of SWEETs in development and stress tolerance in two rose species. Moreover, the cold-response expression patterns of SWEET genes were validated between different cold-resistant rose species, suggesting potential candidates for cold tolerance breeding. These findings lay the groundwork for further functional analysis of the SWEET gene family in roses and contribute to the breeding of cold-tolerant rose varieties.
Article
Biochemical Research Methods
Alma Andersson, Joakim Lundeberg
Summary: In this study, a novel method was proposed to analyze spatial transcriptomics data, which simulated the diffusion of individual transcripts to extract genes with spatial patterns, generating satisfactory results on synthetic and real data. Compared to existing methods, this approach seemed to be less influenced by gene expression levels and showed better time performance when run with multiple cores.
Article
Biochemical Research Methods
Ramon Vinas, Helena Andres-Terre, Pietro Lio, Kevin Bryson
Summary: The study developed a method based on conditional generative adversarial networks to generate realistic transcriptomics data for Escherichia coli and humans. Results showed that the approach performed better in preserving gene expression properties compared to existing simulators, maintaining tissue- and cancer-specific attributes, and exhibiting real gene clusters and ontologies at different scales.
Article
Biochemical Research Methods
Paul Scherer, Maja Trebacz, Nikola Simidjievski, Ramon Vinas, Zohreh Shams, Helena Andres Terre, Mateja Jamnik, Pietro Lio
Summary: Gene expression data is often high dimensional, noisy, and has a low number of samples, making it challenging for learning algorithms. In this article, a method called Gene Interaction Network Constrained Construction (GINCCo) is proposed to construct computational graph models for gene expression data by incorporating the structure of gene interaction networks. The results of a case study on cancer phenotype prediction tasks show that GINCCo outperforms other models while greatly reducing model complexity.
Article
Neurosciences
Gaia Meoni, Leonardo Tenori, Sebastian Schade, Cristina Licari, Chiara Pirazzini, Maria Giulia Bacalini, Paolo Garagnani, Paola Turano, Claudia Trenkwalder, Claudio Franceschi, Brit Mollenhauer, Claudio Luchinat
Summary: Parkinson's disease is the neurological disorder with the highest increase in prevalence. The lack of precise diagnosis at early stages remains a challenge. Metabolomics has provided valuable insights into the molecular basis of PD and potential biomarkers for early detection and treatment efficacy. In this study, NMR was used to analyze serum samples from German PD patients, revealing more pronounced pathological characteristics in male patients and confirming altered levels of acetone and cholesterol. Additionally, stronger oxidative stress markers were detected.
NPJ PARKINSONS DISEASE
(2022)
Article
Multidisciplinary Sciences
Saeed Ahmad, Phasit Charoenkwan, Julian M. W. Quinn, Mohammad Ali Moni, Md Mehedi Hasan, Pietro Lio, Watshara Shoombuatong
Summary: In this study, a computational approach called SCORPION is proposed for the accurate identification of phage virion proteins (PVPs) using only protein primary sequences. By exploring various feature descriptors and machine learning algorithms, optimal baseline models were constructed and a two-step feature selection strategy was used to determine the optimal feature vector. Results demonstrate that SCORPION outperforms existing methods and shows superior predictive performance.
SCIENTIFIC REPORTS
(2022)
Article
Multidisciplinary Sciences
Phasit Charoenkwan, Saeed Ahmed, Chanin Nantasenamat, Julian M. W. Quinn, Mohammad Ali Moni, Pietro Lio, Watshara Shoombuatong
Summary: This study presents a novel meta-predictor, AMYPred-FRL, which utilizes a feature representation learning approach to identify amyloid proteins more accurately. By combining multiple machine learning algorithms and sequence-based feature descriptors, AMYPred-FRL generates 60 probabilistic features and forms a hybrid model. Through cross-validation and independent tests, AMYPred-FRL outperforms existing methods in predictive performance.
SCIENTIFIC REPORTS
(2022)
Article
Computer Science, Information Systems
Xiaofeng Lu, Yuying Liao, Chao Liu, Pietro Lio, Pan Hui
Summary: This article proposes a heterogeneous model fusion federated learning mechanism to address the resource waste issue caused by computing power imbalance in IoT devices. It trains learning models of different scales on devices with lower computing power and evaluates the effectiveness of the method through experiments.
IEEE INTERNET OF THINGS JOURNAL
(2022)
Article
Biology
Phasit Charoenkwan, Nalini Schaduangrat, Mohammad Ali Moni, Pietro Lio, Balachandran Manavalan, Watshara Shoombuatong
Summary: This study presents a novel computational method, SAPPHIRE, for accurately identifying thermophilic proteins (TPPs) using sequence information. The method combines different feature encodings and machine learning algorithms to train baseline models and extract key information of TPPs. SAPPHIRE outperforms existing methods in terms of predictive performance and achieves higher accuracy and correlation coefficient.
COMPUTERS IN BIOLOGY AND MEDICINE
(2022)
Article
Biology
Phasit Charoenkwan, Nalini Schaduangrat, Pietro Lio, Mohammad Ali Moni, Balachandran Manavalan, Watshara Shoombuatong
Summary: This study proposes a novel computational approach, NEPTUNE, for the accurate and large-scale identification of Tumor Homing Peptides (THPs) from sequence information. The results demonstrate that NEPTUNE achieves superior performance in THP prediction and improves interpretability using the SHapley additive explanations method.
COMPUTERS IN BIOLOGY AND MEDICINE
(2022)
Article
Computer Science, Artificial Intelligence
Tiago Azevedo, Alexander Campbell, Rafael Romero-Garcia, Luca Passamonti, Richard A. I. Bethlehem, Pietro Lio, Nicola Toschi
Summary: In this paper, a novel deep neural network architecture is proposed that combines graph neural networks and temporal convolutional networks for learning from both the spatial and temporal components of resting-state functional magnetic resonance imaging (rs-fMRI) data. The model is evaluated using samples from the UK Biobank and Human Connectome Project datasets, showing effectiveness and explainability-related features. This approach lays the groundwork for future deep learning architectures focused on the spatio-temporal nature of rs-fMRI data.
MEDICAL IMAGE ANALYSIS
(2022)
Article
Oncology
Yanan Wang, Yu Guang Wang, Changyuan Hu, Ming Li, Yanan Fan, Nina Otter, Ikuan Sam, Hongquan Gou, Yiqun Hu, Terry Kwok, John Zalcberg, Alex Boussioutas, Roger J. Daly, Guido Montufar, Pietro Lio, Dakang Xu, Geoffrey I. Webb, Jiangning Song
Summary: This study proposes an AI-powered digital staging system that analyzes spatial patterns in the tumor microenvironment to accurately predict survival rates and staging of gastric cancer patients. The results show outstanding model performance and significant improvement over traditional staging systems.
NPJ PRECISION ONCOLOGY
(2022)
Article
Biochemical Research Methods
Pietro Bongini, Franco Scarselli, Monica Bianchini, Giovanna Maria Dimitri, Niccolo Pancino, Pietro Lio
Summary: Drug side-effects have a significant impact on public health, care system costs, and drug discovery processes. Predicting the probability of side-effects before their occurrence is crucial to reduce this impact, especially in drug discovery. By integrating heterogeneous data into a graph dataset, this study successfully utilizes Graph Neural Networks (GNNs) to predict drug side-effects, showing promising results. The experimental results highlight the significance of utilizing relationships between data entities and suggest potential future developments in this field.
IEEE-ACM TRANSACTIONS ON COMPUTATIONAL BIOLOGY AND BIOINFORMATICS
(2023)
Article
Computer Science, Information Systems
Nuruzzaman Faruqui, Mohammad Abu Yousuf, Md Whaiduzzaman, A. K. M. Azad, Salem A. Alyami, Pietro Lio, Muhammad Ashad Kabir, Mohammad Ali Moni
Summary: The Internet of Medical Things (IoMT) has become an attractive target for cybercriminals due to its market value and rapid growth. However, IoMT devices have limited computational capabilities, making them vulnerable to cyber-attacks. To address this, a novel Intrusion Detection System (IDS) called SafetyMed is proposed, which combines Convolutional Neural Networks (CNN) and Long Short-Term Memory (LSTM) networks to defend against intrusion from sequential and grid data. SafetyMed has shown high detection rates and accuracy, making it a potential game-changer in vulnerable sectors like the medical industry.
Article
Health Care Sciences & Services
Md. Martuza Ahamad, Sakifa Aktar, Md. Jamal Uddin, Md. Rashed-Al-Mahfuz, A. K. M. Azad, Shahadat Uddin, Salem A. Alyami, Iqbal H. Sarker, Asaduzzaman Khan, Pietro Lio, Julian M. W. Quinn, Mohammad Ali Moni
Summary: Good vaccine safety and reliability are crucial for countering infectious diseases effectively. This study aims to reduce adverse reactions to COVID-19 vaccines by identifying common factors through patient data analysis and classification. Patient medical histories and postvaccination effects were examined, and statistical and machine learning approaches were used. The analysis revealed that prior illnesses, hospital admissions, and SARS-CoV-2 reinfection were significantly associated with poor patient reactions.
Proceedings Paper
Acoustics
Alexander Campbell, Lorena Qendro, Pietro Lio, Cecilia Mascolo
Summary: This article proposes an approach for estimating predictive uncertainty using early exit ensembles. Empirical evaluation shows that this method performs well in terms of accuracy and uncertainty metrics, while also providing significant computational speed-up and memory reduction compared to single model baselines.
2022 IEEE INTERNATIONAL CONFERENCE ON ACOUSTICS, SPEECH AND SIGNAL PROCESSING (ICASSP)
(2022)
Review
Biology
Phasit Charoenkwan, Nalini Schaduangrat, Md Mehedi Hasan, Mohammad Ali Moni, Pietro Lio, Watshara Shoombuatong
Summary: This article comprehensively investigates 14 state-of-the-art TPP predictors and summarizes their characteristics and advantages and disadvantages. Through comparative analysis, it provides future perspectives for the development of more accurate and efficient TPP predictors.