Article
Biochemical Research Methods
Wen-Ya Zhang, Junhai Xu, Jun Wang, Yuan-Ke Zhou, Wei Chen, Pu-Feng Du
Summary: With the advancement of high-throughput sequencing technology, genomic sequences have exponentially increased, leading to the introduction of machine learning methods for genome annotation and analysis. To facilitate the study of genomic sequences, the KNIndex database was developed to deposit and visualize physicochemical properties of k-tuple nucleotides, providing a user-friendly interface for browsing, querying, visualizing, and downloading these properties.
BRIEFINGS IN BIOINFORMATICS
(2021)
Article
Engineering, Electrical & Electronic
S. Samsudeen, G. Senthil Kumar
Summary: In this study, a novel federated unsupervised learning-based predictive model (FeduLPM) technique is proposed to control the speed of customizable automotive variants by analyzing accident data from different locations. The method distributes accident data to local models and aggregates trained parameters to generate a global model, providing accurate speed limit suggestions. Experimental results show that the proposed FeduLPM achieved 96.7% accuracy in processing data from various locations in Bengaluru, making it a better solution to prevent accidents for bike and car drivers.
IEEE SENSORS JOURNAL
(2023)
Article
Biology
Diana Rios-Szwed, Vanesa Alvarez, Luis Sanchez-Pulido, Elisa Garcia-Wilson, Hao Jiang, Susanne Bandau, Angus Lamond, Constance Alabert
Summary: FAM111A is a replisome-associated protein that plays an important role in the activation of DNA replication origins. Dominant mutations within its trypsin-like peptidase domain are linked to severe human developmental syndrome. Research has shown that under normal conditions, FAM111A promotes DNA replication, but in a disease context, its unrestrained expression can lead to DNA damage and cell death.
LIFE SCIENCE ALLIANCE
(2023)
Article
Biochemistry & Molecular Biology
Akanksha Rajput, Manoj Kumar
Summary: The Ebola virus, a deadly pathogen since 1976, has led researchers to develop computational models for drug discovery. Using molecular descriptors, a predictive model for anti-Ebola compounds was developed and integrated into a web server for scientific use.
MOLECULAR DIVERSITY
(2022)
Article
Environmental Sciences
Pengliang Wei, Ran Huang, Tao Lin, Jingfeng Huang
Summary: This study introduces a workflow that utilizes a deep semantic segmentation model to extract rice distribution information in regions with limited training samples. By training the model on pseudo-labels, the time-consuming annotation process for ground truth data can be reduced. Experimental results show that the proposed method outperforms existing approaches in terms of accurately extracting rice area and spatial distribution information.
Article
Food Science & Technology
Eirini Pegiou, Roland Mumm, Robert D. Hall
Summary: A study found that the chemical composition of asparagus changes during cooking, with some substances increasing while others decreasing. Profiles of asparagus metabolites were analyzed using GC-MS and LC-MS, and the flavor attributes were evaluated by a taste panel. The study revealed the key biochemical pathways and chemical transformations relevant to asparagus flavor.
LWT-FOOD SCIENCE AND TECHNOLOGY
(2023)
Article
Multidisciplinary Sciences
Thorsten Mosler, Francesca Conte, Gabriel M. C. Longo, Ivan Mikicic, Nastasja Kreim, Martin M. Moeckel, Giuseppe Petrosino, Johanna Flach, Joan Barau, Brian Luke, Vassilis Roukos, Petra Beli
Summary: The study reveals that DDX41 plays a crucial role in regulating the accumulation of R-loop and double strand DNA breaks in gene promoters through mapping the R-loop proximal proteome in human cells.
NATURE COMMUNICATIONS
(2021)
Review
Multidisciplinary Sciences
Chunyan Ao, Shihu Jiao, Yansu Wang, Liang Yu, Quan Zou
Summary: The rapid growth of biological sequences has driven the application of machine learning in this field, focusing on function and modification classification. Establishing a support website to provide information and datasets for classification methods, discussing current challenges and future prospects.
Article
Multidisciplinary Sciences
Luis Bermudez-Guzman
Summary: This study demonstrates the existence and importance of non-oncogenic addiction to DNA repair in cancer, which may assist in identifying prognostic biomarkers and therapeutic opportunities.
SCIENTIFIC REPORTS
(2021)
Article
Biochemistry & Molecular Biology
Commodore P. St Germain, Hongchang Zhao, Vrishti Sinha, Lionel A. Sanz, Frederic Chedin, Jacqueline H. Barlow
Summary: Conflicts between transcription and replication machinery can lead to replication stress and genome instability. The newly developed TRIPn-Seq method allows the identification of genomic loci prone to transcription-replication interactions. Using TRIPn-Seq, the authors mapped 1009 unique transcription-replication interactions in mouse primary B cells, which were enriched at transcription start sites and early replicating regions.
NUCLEIC ACIDS RESEARCH
(2022)
Article
Microbiology
Yan Lin, Meili Sun, Junjie Zhang, Mingyan Li, Keli Yang, Chengyan Wu, Hasan Zulfiqar, Hongyan Lai
Summary: This study aims to develop a machine learning-based model for predicting promoters in Klebsiella aerogenes. The model utilizes a unique encoding and optimization method to accurately identify promoter sequences in K. aerogenes.
FRONTIERS IN MICROBIOLOGY
(2023)
Article
Chemistry, Medicinal
Jun Hu, Wen-Wu Zeng, Ning-Xin Jia, Muhammad Arif, Dong-Jun Yu, Gui-Jun Zhang
Summary: A new sequence feature extraction strategy called TPSO is developed for predicting DNA-binding proteins (DBPs), which achieves higher accuracy and Matthew's correlation coefficient value compared to existing methods. The TPSO-DBP method utilizes TPSO and a deep learning framework to learn the relationship between input features and DBPs.
JOURNAL OF CHEMICAL INFORMATION AND MODELING
(2023)
Article
Chemistry, Medicinal
Jun Hu, Wen-Wu Zeng, Ning-Xin Jia, Muhammad Arif, Dong-Jun Yu, Gui-Jun Zhang
Summary: A new three-part sequence order feature extraction strategy (TPSO) is developed to predict DNA-binding protein (DBP) by extracting more discriminative information from protein sequences. A deep learning-based method called TPSO-DBP is proposed, which achieves an accuracy of 87.01% and a significantly higher Matthews correlation coefficient value (0.741) compared to existing DBP prediction methods.
JOURNAL OF CHEMICAL INFORMATION AND MODELING
(2023)
Article
Chemistry, Applied
Quynh Phan, Elizabeth Tomasino
Summary: This study utilized an advanced lipidomic profiling approach to analyze commercial Pinot noir wines, revealing that wine lipids have a strong potential for classifying wines by origin, with the top 58 lipids playing a significant role in discrimination.
Article
Environmental Sciences
Xinxin Guo, Chaoying Zhao, Guangrong Li, Mimi Peng, Qin Zhang, Pierluigi Confuorto, Federico Raspini, Matteo Del Soldato, Chiara Cappadonia, Simon Plank, Mariano Di Napoli
Summary: Synthetic Aperture Radar Interferometry (InSAR) is an effective technique for monitoring large-scale ground deformation with high spatial resolution. However, it is challenging to obtain a spatially continuous deformation map due to SAR decorrelation or distortion. In this study, we propose a multifactor-based machine learning model called the K-RFR model, which combines K-means clustering and random forest regression algorithms to reconstruct a continuous deformation map. The model takes into account various influence factors on ground deformation, such as land use, geological engineering, and groundwater extraction. The study conducted in Xi'an, China, using the SBAS-InSAR technique, demonstrates the effectiveness of the proposed model in predicting ground deformation. The new model outperforms traditional interpolation methods, achieving a higher correlation coefficient with the InSAR measurements.
Review
Biochemical Research Methods
Zhourun Wu, Qing Liao, Bin Liu
BRIEFINGS IN BIOINFORMATICS
(2020)
Article
Biochemical Research Methods
Chen-Chen Li, Bin Liu
BRIEFINGS IN BIOINFORMATICS
(2020)
Article
Biochemical Research Methods
Bin Liu, Chen-Chen Li, Ke Yan
BRIEFINGS IN BIOINFORMATICS
(2020)
Article
Biochemical Research Methods
Bin Liu, Yulin Zhu, Ke Yan
BRIEFINGS IN BIOINFORMATICS
(2020)
Article
Medicine, Research & Experimental
Bin Liu, Zhihua Luo, Juan He
MOLECULAR THERAPY-NUCLEIC ACIDS
(2020)
Article
Biochemical Research Methods
Yi-Jun Tang, Yi-He Pang, Bin Liu
Article
Biochemical Research Methods
Jiangyi Shao, Ke Yan, Bin Liu
Summary: The FoldRec-C2C predictor globally incorporates protein interactions for protein fold recognition, treating it as an information retrieval task in natural language processing. Tested on the LINDAHL dataset, FoldRec-C2C outperforms 34 state-of-the-art methods in the field.
BRIEFINGS IN BIOINFORMATICS
(2021)
Article
Biochemical Research Methods
Jiangyi Shao, Bin Liu
Summary: This study introduces a network-based predictor ProtFold-DFG for protein fold recognition, utilizing Directed Fusion Graph (DFG), KL divergence, and PageRank algorithm to enhance recognition accuracy. Experimental results demonstrate that ProtFold-DFG outperforms 35 other methods on the LINDAHL dataset, making it a promising approach for protein fold recognition.
BRIEFINGS IN BIOINFORMATICS
(2021)
Article
Biology
Hang Wei, Yuxin Ding, Bin Liu
COMPUTATIONAL BIOLOGY AND CHEMISTRY
(2020)
Article
Biochemical Research Methods
Ke Yan, Jie Wen, Jin-Xing Liu, Yong Xu, Bin Liu
Summary: The study proposed two novel algorithms, TSVM-fold and ESVM-fold, utilizing sequence similarity scores generated by multiple template-based methods for protein fold recognition prediction. Experimental results showed that these algorithms outperform some state-of-the-art methods in rigorous benchmark datasets.
IEEE-ACM TRANSACTIONS ON COMPUTATIONAL BIOLOGY AND BIOINFORMATICS
(2021)
Article
Biochemical Research Methods
Hang Wei, Qing Liao, Bin Liu
Summary: Identifying lncRNA-disease associations is crucial for exploring disease mechanisms and molecular drug discovery. However, current fusion strategies fail to remove noisy and irrelevant information, leading to low predictive performance. iLncRNAdis-FB proposes a new computational predictor based on CNN to integrate feature blocks from different data sources, achieving better prediction accuracy.
IEEE-ACM TRANSACTIONS ON COMPUTATIONAL BIOLOGY AND BIOINFORMATICS
(2021)
Article
Biochemical Research Methods
Yumeng Liu, Xiaolong Wang, Bin Liu
Summary: Intrinsically disordered proteins/regions (IDPs/IDRs) are important for biological functions, and accurate prediction is crucial for protein structure and function predictions. However, most existing methods tend to predict fully ordered proteins as disordered, ignoring the fact that most newly sequenced proteins are fully ordered. The proposed RFPR-IDP method, trained on both ordered and disordered proteins, outperforms existing predictors in predicting both ordered and disordered proteins.
BRIEFINGS IN BIOINFORMATICS
(2021)
Article
Biochemical Research Methods
Zhourun Wu, Qing Liao, Bin Liu
Summary: Protein complexes are key units for studying a cell system, and high-throughput approaches have enabled the determination of PPI data. The proposed mutual important interacting partner relation and the new algorithm idenPC-MIIP show improved performance in identifying protein complexes compared to existing methods.
BRIEFINGS IN BIOINFORMATICS
(2021)
Article
Computer Science, Artificial Intelligence
Ke Yan, Jie Wen, Yong Xu, Bin Liu
Summary: Protein fold recognition is crucial for understanding protein functions and drug design. New methods (MVLR and MLDH-Fold) were proposed to improve predictive performance by combining different views of protein sequences. Experimental results show that these computational methods outperform other predictors, indicating their usefulness for protein fold recognition.
Article
Biochemical Research Methods
Bin Liu, Shuangyan Jiang, Quan Zou
BRIEFINGS IN BIOINFORMATICS
(2020)