Article
Biochemical Research Methods
Katie Ovens, Farhad Maleki, B. Frank Eames, Ian McQuillan
Summary: This paper introduces a tool called Juxtapose and similarity measures for comparative transcriptomics between organisms. The tool utilizes word embedding strategy to generate gene embeddings and was evaluated based on its ability to embed nodes consistently and generate biologically informative results. Juxtapose is shown to globally align synthesized networks and identify conserved areas in real gene co-expression networks, providing a measure of similarity based on cosine distance between GCN embeddings.
BMC BIOINFORMATICS
(2021)
Article
Chemistry, Multidisciplinary
Ibrahim Riza Hallac, Betul Ay, Galip Aydin
Summary: This study presented a novel doc2vec-based representation method for high-quality social media user representations and conducted various experiments to investigate the performance of different text representation techniques. Additionally, a new social media dataset was shared for the study of social media user data.
APPLIED SCIENCES-BASEL
(2021)
Article
Computer Science, Information Systems
Chenkai Guo, Dengrong Huang, Naipeng Dong, Jianwen Zhang, Jing Xu
Summary: Although many embedding approaches have been proposed for code representation of mobile applications, insufficient attention has been paid to the event-driven nature of their running. This paper introduces a callback-based hierarchical embedding method Callback2Vec, which captures the running behavior of callbacks through a fine-grained callback sequence generation algorithm.
INFORMATION SCIENCES
(2021)
Article
Biochemical Research Methods
Guiying Wu, Xiangyu Li, Wenbo Guo, Zheng Wei, Tao Hu, Yiran Shan, Jin Gu
Summary: The inference of gene co-expression associations is a fundamental task in large-scale transcriptomic data analysis. We proposed a method called JEBIN that integrates multiple expression datasets to learn a low-dimensional consensus representation for genes. Simulation experiments confirmed the effectiveness and scalability of JEBIN, and it outperformed commonly used integration methods. JEBIN was also applied to study the gene co-expression patterns of hepatocellular carcinoma (HCC) and identified differentially co-expressed ligand-receptor pairs.
BRIEFINGS IN BIOINFORMATICS
(2022)
Article
Computer Science, Information Systems
Jinyuan Fang, Shangsong Liang, Zaiqiao Meng, Maarten De Rijke
Summary: Network embedding is an important technique for learning effective representations of network data. However, existing embedding methods disregard the spherical nature of input features, which may lead to less effective representations. To address this issue, this paper proposes a hyperspherical variational co-embedding method (HCAN) for attributed networks, which captures the correlations between nodes and attributes by learning their representations in a unified hyperspherical space.
ACM TRANSACTIONS ON INFORMATION SYSTEMS
(2022)
Review
Pharmacology & Pharmacy
Bing-Xue Du, Yuan Qin, Yan-Feng Jiang, Yi Xu, Siu-Ming Yiu, Hui Yu, Jian-Yu Shi
Summary: This review comprehensively investigates DL-based CPI prediction, covering popular databases, representation methods, and state-of-the-art models. Current challenges and trends suggest that crucial progress lies in better CPI prediction and key approaches in practical applications.
DRUG DISCOVERY TODAY
(2022)
Article
Biology
You Li, Jianyi Lyu, Yaoqun Wu, Yuewu Liu, Guohua Huang
Summary: RNA-protein interactions are crucial in biological processes and their aberration is associated with human diseases. A sequence semantics-based method, PRIP, is introduced to predict RNA-binding interfaces and explore the mechanism of RNA-protein interactions from a semantics point of view.
Article
Immunology
Miri Ostrovsky-Berman, Boaz Frankel, Pazit Polak, Gur Yaari
Summary: The adaptive branch of the immune system learns and remembers pathogenic patterns through T- and B-cell receptors, posing challenges in extracting meaningful biological information. Immune2vec, a natural language processing-based embedding technique, is validated for BCR repertoire sequencing data, demonstrating its reliability in preserving immune sequencing data information for feature extraction and exploratory analysis, showing promise in stratifying distinct clinical conditions.
FRONTIERS IN IMMUNOLOGY
(2021)
Article
Engineering, Industrial
Koutarou Yamashita, Fumiyo Ito, Kyosuke Hasumoto, Masayuki Goto
Summary: Recently, a large number of cooking recipes have been posted and shared on the Internet. Various machine learning techniques have been proposed to analyze these recipes, calculating distances and evaluating similarities using constructed semantic spaces. Additionally, a method to analyze the diversity of recipes has been proposed and applied to an actual recipe site.
INDUSTRIAL ENGINEERING AND MANAGEMENT SYSTEMS
(2023)
Article
Computer Science, Artificial Intelligence
I-Chung Hsieh, Cheng-Te Li
Summary: Attributed Network Embedding (ANE) is a method that learns low-dimensional vectors to preserve both the network structure and node attributes in the embedding space. In this paper, we propose a novel ANE model called Context Co-occurrence-aware Attributed Network Embedding (CoANE), which models the context attributes and uses convolutional mechanism to encode positional information. The learning of context co-occurrence can capture the latent social circles of each node. Experimental results show that CoANE outperforms state-of-the-art ANE models significantly.
IEEE TRANSACTIONS ON KNOWLEDGE AND DATA ENGINEERING
(2023)
Article
Genetics & Heredity
Xinfeng Wang, Akhilesh K. Bajpai, Qingqing Gu, David G. Ashbrook, Athena Starlard-Davenport, Lu Lu
Summary: In this study, a cross-species integrative approach was used to identify genes strongly associated with acute myeloid leukemia (AML). Through the integration of protein-protein interactions, transcription factors, gene function, genetic regulation, and coding sequence variants, key hub genes in AML were identified. The findings highlight the importance of this cross-species approach in identifying multiple key candidate genes in AML.
FRONTIERS IN GENETICS
(2023)
Article
Computer Science, Artificial Intelligence
Carlos Perinan-Pascual
Summary: The study focused on developing a model to automatically estimate the semantic relatedness between words, showing that the model's performance depends on independently constructed word embeddings and their interaction. By using a weighted average of cosine-similarity coefficients derived from independent word embeddings in a double vector space, high correlations with human judgments were achieved. The evaluation of word associations through a measure that considers both the rank ordering of word pairs and the strength of associations revealed findings unnoticed by traditional measures like Spearman's and Pearson's correlation coefficients.
ARTIFICIAL INTELLIGENCE REVIEW
(2022)
Article
Computer Science, Artificial Intelligence
Piotr Bielak, Kamil Tagowski, Maciej Falkiewicz, Tomasz Kajdanowicz, Nitesh V. Chawla
Summary: This paper discusses the challenges of representation learning on dynamic graphs and proposes a framework called FILDNE for incorporating advances in static representation learning into dynamic graphs. FILDNE reduces computational costs while improving quality measure gains by applying static representation learning methods to dynamic graphs.
KNOWLEDGE-BASED SYSTEMS
(2022)
Article
Multidisciplinary Sciences
Linhua Wang, Mirjana Maletic-Savatic, Zhandong Liu
Summary: Spatially resolved transcriptomics is a new technique for mapping transcriptional information within tissues. In this study, the authors present MIST, a computational tool that detects molecular regions and denoises missing gene expression values using region-specific imputation.
NATURE COMMUNICATIONS
(2022)
Article
Multidisciplinary Sciences
Chang Su, Zichun Xu, Xinning Shan, Biao Cai, Hongyu Zhao, Jingfei Zhang
Summary: The advancement of scRNA-seq technology enables the direct inference of co-expressions in specific cell types, but existing methods fail to address the challenges of sequencing depth variations and measurement errors. CS-CORE is a statistical approach that accurately estimates and tests cell-type-specific co-expressions while considering these challenges. Evaluations demonstrate that CS-CORE outperforms existing methods in terms of accuracy and identification of relevant co-expressions. Applied to scRNA-seq data from Alzheimer's disease and COVID-19 patients, CS-CORE identifies reproducible and biologically relevant cell-type-specific co-expressions and differential co-expressions.
NATURE COMMUNICATIONS
(2023)
Review
Immunology
Yang Xiang, Jingcheng Du, Kayo Fujimoto, Fang Li, John Schneider, Cui Tao
Summary: The goal to end the HIV epidemic presents various challenges, but artificial intelligence has shown great potential in developing effective prevention intervention strategies.
Article
Immunology
Andrew T. Lian, Jingcheng Du, Lu Tang
Summary: This study utilized machine learning and natural language processing to identify COVID-19 vaccine adverse events from Twitter data. The research found that the four most populous states in the US were the areas with the most discussions about adverse events on Twitter, and the most common adverse effects were sore to touch, fatigue, and headache. The findings demonstrate the feasibility of using social media data to monitor vaccine adverse events.
Article
Biochemistry & Molecular Biology
Andi Liu, Astrid M. Manuel, Yulin Dai, Brisa S. Fernandes, Nitesh Enduru, Peilin Jia, Zhongming Zhao
Summary: This study used large-scale proteomic datasets and GWAS data to investigate the molecular pathways and potential drug targets specific to different brain regions in Alzheimer's disease. By applying a network-based tool, the researchers identified specific module genes associated with AD and pinpointed three potential drug targets in the parahippocampal gyrus.
HUMAN MOLECULAR GENETICS
(2022)
Article
Engineering, Chemical
Huizhen Shen, Li Chen, Cailong Zhou, Jingcheng Du, Chenyang Lu, Hao Yang, Luxi Tan, Xinjuan Zeng, Lichun Dong
Summary: In this study, NZVI was immobilized on porous TpPa-1 covalent organic frameworks (COFs) using dopamine as a connector, resulting in the formation of Fe-0/TpPa-1@DOPA composite. The composite exhibited excellent performance in Cr(VI) removal, making it a promising material for wastewater treatment.
SEPARATION AND PURIFICATION TECHNOLOGY
(2022)
Article
Genetics & Heredity
Guangsheng Pei, Fangfang Yan, Lukas M. Simon, Yulin Dai, Peilin Jia, Zhongming Zhao
Summary: Single-cell RNA sequencing (scRNA-seq) is revolutionizing the study of complex and dynamic cellular mechanisms. In this study, the researchers present deCS, an automatic cell type annotation method that enhances annotation accuracy using comprehensive human cell type expression profiles and marker genes. The results show that expanding the references is critical for improving annotation accuracy, and deCS significantly reduces computation time and increases accuracy compared to existing tools. The researchers also demonstrate the broad utility of deCS in identifying trait-cell type associations in human complex traits, providing deep insights into disease pathogenesis.
GENOMICS PROTEOMICS & BIOINFORMATICS
(2023)
Article
Oncology
Erika Y. Faraoni, Kanchan Singh, Vidhi Chandra, Olivereen Le Roux, Yulin Dai, Ismet Sahin, Baylee J. O'Brien, Lincoln N. Strickland, Le Li, Emily Vucic, Amanda N. Warner, Melissa Pruski, Trent Clark, George Van Buren, Nirav C. Thosani, John S. Bynon, Curtis J. Wray, Dafna Bar-Sagi, Kyle L. Poulsen, Lana A. Vornik, Michelle I. Savage, Shizuko Sei, Altaf Mohammed, Zhongming Zhao, Powel H. Brown, Tingting Mills, Holger K. Eltzschig, Florencia McAllister, Jennifer M. Bailey-Lundberg
Summary: The microenvironment of PDAC is desmoplastic and immunosuppressive. CD73 is overexpressed in the tumor microenvironment and could be a target for immunotherapy.
Article
Biochemistry & Molecular Biology
Jiajinlong Kang, Yulin Dai, Jinze Li, Huihui Fan, Zhongming Zhao
Summary: This article reports the first single-cell level analysis of eccDNA in glioblastoma (GBM) samples. The study revealed the presence of potential mobile enhancers acting in a trans-regulation manner in GBM. This research provides insights into the novel features of eccDNA in the cellular context of brain tumors, highlighting the importance of investigating eccDNA at the single-cell level.
COMPUTATIONAL AND STRUCTURAL BIOTECHNOLOGY JOURNAL
(2023)
Article
Genetics & Heredity
Toshiyuki Itai, Peilin Jia, Yulin Dai, Jingchun Chen, Xiangning Chen, Zhongming Zhao
Summary: Investigating functional, temporal, and cell-type expression features of mutations is important for understanding schizophrenia. This study collected and analyzed mutations in schizophrenia patients and identified genes that are neurologically important and have specific expression patterns during prenatal development. Results suggest that gene expression patterns in specific cell types during early fetal stages may impact the risk of schizophrenia in adulthood.
AMERICAN JOURNAL OF MEDICAL GENETICS PART B-NEUROPSYCHIATRIC GENETICS
(2023)
Article
Materials Science, Multidisciplinary
Ziye Song, Qian Sun, Jingcheng Du, Linghao Liu, Wen He, Yiming Xu, Jiangtao Liu
Summary: Covalent organic frameworks (COFs) are emerging as a promising class of porous materials with high porosity, low density, and excellent physicochemical stability. Smart COFs with special structures and functional groups have attracted considerable attention due to their ability to respond to external stimuli. However, fabricating smart COF membranes with adjustable pore sizes for gradient separation of organic pollutants remains a challenge.
ACS APPLIED POLYMER MATERIALS
(2023)
Article
Engineering, Chemical
Qian Sun, Jingcheng Du, Linghao Wang, Ayan Yao, Ziye Song, Linghao Liu, Dong Cao, Ji Ma, Weiwang Lim, Wen He, Shabi Ul Hassan, Cailong Zhou, Jiangtao Liu
Summary: In this study, a COF membrane with unique super-wettability was prepared using collected COF nanofibers by filtration assembly method. It showed controllable and switchable separation performance in oil/water separation field, achieving high separation efficiency and permeance.
SEPARATION AND PURIFICATION TECHNOLOGY
(2023)
Article
Materials Science, Multidisciplinary
Ziye Song, Qian Sun, Jingcheng Du, Linghao Liu, Wen He, Yiming Xu, Jiangtao Liu
Summary: Covalent organic frameworks (COFs) are emerging crystalline porous materials with desirable properties. Smart COFs with special structures and functional groups can respond to external stimuli. However, fabricating smart COF membranes with adjustable pore size for gradient separation of organic pollutants remains a challenge.
ACS APPLIED POLYMER MATERIALS
(2023)
Article
Chemistry, Multidisciplinary
Wen He, Jingcheng Du, Linghao Liu, Qian Sun, Ziye Song, Ji Ma, Dong Cao, Weiwang Lim, Shabi Ul Hassan, Jiangtao Liu
Summary: Graphene oxide (GO) tuned polyimide carbon molecular sieve (CMS) membranes were prepared by carbonization, showing high permeability, selectivity, and stability. The gas sorption capability increased with the carbonization temperature, creating more micropores under higher temperatures under GO guidance. GO guidance and subsequent carbonization enhanced H-2 permeability and selectivity, surpassing state-of-the-art materials. The CMS membranes transitioned from a polymeric structure to a denser graphite structure with increasing carbonization temperature, achieving ultrahigh selectivities for various gas pairs while maintaining moderate H-2 gas permeabilities.
Proceedings Paper
Computer Science, Artificial Intelligence
Shiqiang Tao, Wei-Chun Chou, Jianfu Li, Jingcheng Du, Pritham Ram, Rashmie Abeysinghe, Hua Xu, Xiaoqian Jiang, Peter W. Rose, Lucila Ohno-Machado, Guo-Qiang Zhang
Summary: The National Institute of Health (NIH) has launched the RADx Radical research collaboratives (RADx-rad) to advance new, non-traditional approaches for COVID-19 testing. They have developed the web application IMI-CDE to facilitate the mapping of study variables to common data elements (CDEs) for researchers, increasing data interoperability. The application has been piloted with positive feedback from beta-testers.
2022 IEEE 10TH INTERNATIONAL CONFERENCE ON HEALTHCARE INFORMATICS (ICHI 2022)
(2022)
Proceedings Paper
Computer Science, Artificial Intelligence
Jingqi Wang, Yaoyun Zhang, Bin Lin, Huy Anh Pham, Long He, Jingcheng Du, Frank Manion
Summary: The creation of high-quality annotated corpora is crucial for the development of machine and deep-learning models for Information Extraction and Natural Language Processing. This paper presents LANN, a text annotation tool that supports team-based annotation, quality controls, and machine learning assistance throughout the annotation project workflow.
2022 IEEE 10TH INTERNATIONAL CONFERENCE ON HEALTHCARE INFORMATICS (ICHI 2022)
(2022)
Article
Mathematical & Computational Biology
Qingyu Chen, Alexis Allot, Robert Leaman, Rezarta Islamaj, Jingcheng Du, Li Fang, Kai Wang, Shuo Xu, Yuefu Zhang, Parsa Bagherzadeh, Sabine Bergler, Aakash Bhatnagar, Nidhir Bhavsar, Yung-Chun Chang, Sheng-Jie Lin, Wentai Tang, Hongtong Zhang, Ilija Tavchioski, Senja Pollak, Shubo Tian, Jinfeng Zhang, Yulia Otmakhova, Antonio Jimeno Yepes, Hang Dong, Honghan Wu, Richard Dufour, Yanis Labrak, Niladri Chatterjee, Kushagri Tandon, Frejus A. A. Laleye, Loic Rakotoson, Emmanuele Chersoni, Jinghang Gu, Annemarie Friedrich, Subhash Chandra Pujari, Mariia Chizhikova, Naveen Sivadasan, V. G. Saipradeep, Zhiyong Lu
Summary: The COVID-19 pandemic has had a severe impact on global society, leading to a rapid growth in related literature. To address the challenges of manual curation and interpretation, the BioCreative LitCovid track called for a community effort to automate topic annotation. Nineteen teams participated, achieving higher scores compared to existing methods.
DATABASE-THE JOURNAL OF BIOLOGICAL DATABASES AND CURATION
(2022)
