Article
Environmental Sciences
Ona Torra, Marcel Hurlimann, Carol Puig-Polo, Mariano Moreno-de-las-Heras
Summary: This research assesses badland susceptibility and its governing factors at different spatial scales using the random forest approach. The feasibility of upscaling a RF model trained in a small area to a large region has been evaluated and confirmed. The final modelling results have facilitated the development of a Badland Susceptibility Map for the full extension of Catalonia with high predictive accuracy, emphasizing lithology and NDVI as the main conditioning factors.
ENVIRONMENTAL RESEARCH
(2023)
Article
Forestry
James J. Worrall, Gerald E. Rehfeldt
Summary: Research indicates that future climate change will have extensive impacts on 14 upland tree species in southwestern Colorado, with the majority of habitats being lost or threatened, and three species potentially locally extirpated. However, strategies have been proposed to utilize these projections for effective management, assisting species migration, favoring future-adapted species, and identifying potential climate refugia.
Article
Biochemistry & Molecular Biology
Angelica M. Walker, Ashley Cliff, Jonathon Romero, Manesh B. Shah, Piet Jones, Joao Gabriel Felipe Machado Gazolla, Daniel A. Jacobson, David Kainer
Summary: Gene-to-gene networks are important tools for studying relationships between genes. Random Forest and iterative Random Forest methods can produce high-quality gene-to-gene networks. The study validates the use of synthetic and empirical data to compare these methods.
COMPUTATIONAL AND STRUCTURAL BIOTECHNOLOGY JOURNAL
(2022)
Article
Chemistry, Medicinal
Robert P. Sheridan
Summary: Similar predictions are observed for the majority of molecules across different versions of the random forest models for ADMET end points. However, a small minority of molecules show substantial shifts in predictions over a few versions. These shifting molecules tend to have more accurate predictions in later versions. This Perspective investigates metrics to identify and predict substantial shifts in molecule predictions.
JOURNAL OF CHEMICAL INFORMATION AND MODELING
(2022)
Article
Environmental Sciences
Li Huang, Feiyue Mao, Lin Zang, Yunquan Zhang, Yi Zhang, Taixin Zhang
Summary: This study investigated PM1 concentrations in central and eastern China using a random forest model, revealing higher levels of PM1 pollution in the Beijing-Tianjin-Hebei region. Approximately 75% of the population was found to be living in an environment with PM1 concentrations exceeding 35μg/m³.
ENVIRONMENTAL POLLUTION
(2021)
Article
Geosciences, Multidisciplinary
Guangsheng Zhou, Hongrui Ren, Tong Liu, Li Zhou, Yuhe Ji, Xingyang Song, Xiaomin Lv
Summary: Understanding the impact of climate change on vegetation and its evolution trend requires accurate and long-term regional vegetation data. This study developed a new method for mapping regional vegetation using terrain-climate-remote sensing. The method, supported by Google Earth Engine and the random forest algorithm, utilized information from remote sensing, climate, and terrain to achieve a 10-meter spatial resolution vegetation map of the Qinghai-Xizang Plateau in 2020.
SCIENCE CHINA-EARTH SCIENCES
(2023)
Article
Agriculture, Multidisciplinary
Jashanjeet Kaur Dhaliwal, Dinesh Panday, Debasish Saha, Jaehoon Lee, Sindhu Jagadamma, Sean Schaeffer, Alemu Mengistu
Summary: Accurate prediction of crop yield is crucial for effective agricultural management decisions. Machine learning methods have gained attention in crop yield prediction. This study evaluated six machine learning algorithms and identified key determinants of cotton yield. The results showed that tree-based models performed well in predicting cotton yield.
COMPUTERS AND ELECTRONICS IN AGRICULTURE
(2022)
Article
Geosciences, Multidisciplinary
Kamal Rana, Ugur Ozturk, Nishant Malik
Summary: This study proposes a method to classify the triggering mechanisms of landslides in existing inventories, enabling their use in landslide hazard modeling by utilizing machine-learning classifier random forest with geometric characteristics as feature space. Application of the method to six landslide inventories showed high accuracy, and it was demonstrated that landslides with identical trigger mechanisms exhibit similar geometric properties.
GEOPHYSICAL RESEARCH LETTERS
(2021)
Article
Computer Science, Information Systems
Md Nasim Adnan, Ryan H. L. Ip, Michael Bewong, Md Zahidul Islam
Summary: The proposed decision forest algorithm in this paper achieves better balance through effective synchronization of diversity from different sources, leading to significant improvement in accuracy according to empirical evaluations. It is also competitive in terms of complexity and other relevant parameters.
INFORMATION SCIENCES
(2021)
Article
Environmental Sciences
Ryan Paulik, Conrad Zorn, Liam Wotherspoon
Summary: Flood damage is the result of complex interactions between flooding processes and socio-economic elements. Damage assessments for residential buildings rely on a modelled representation of local damage factors. Multivariable model approaches using flood hazard and building characteristics can improve damage prediction, but model transfer between geographical contexts can lead to performance loss.
JOURNAL OF FLOOD RISK MANAGEMENT
(2023)
Article
Computer Science, Artificial Intelligence
Li Lang, Liang Tiancai, Ai Shan, Tang Xiangyan
Summary: An improved random forest (RF) algorithm is proposed for wind pressure prediction in this paper, which shows good performance in predicting wind pressure coefficients of high-rise buildings and maintaining acceptable relative error in engineering terms.
INTERNATIONAL JOURNAL OF INTELLIGENT SYSTEMS
(2021)
Article
Engineering, Chemical
Binglong Zhao, Shunxuan Hu, Xuemin Zhao, Baonan Zhou, Junguo Li, Wei Huang, Guohua Chen, Changning Wu, Ke Liu
Summary: In this study, machine learning models were used to simulate the migration behavior of minerals during coal slime flotation. The random forest model showed the highest accuracy and avoided the need for retraining. Particle size and composition were found to play a significant role in coal slime flotation.
ADVANCED POWDER TECHNOLOGY
(2022)
Article
Nanoscience & Nanotechnology
Kun Wang, Jinggeng Gao, Hu Li, Li Liang
Summary: In this paper, a novel method called RF-TStacking is proposed for short-term load forecasting. The importance of influencing factors of the load is estimated using random forest, and short-term load forecasting is achieved through the integration of LightGBM and random forest. Bayesian optimization is used to improve the accuracy of the selection of influencing factors. Testing on load data from a region in northwest China shows that the model provides stable prediction results.
Article
Mathematical & Computational Biology
Alena Orlenko, Jason H. Moore
Summary: The study compared different methods for feature importance estimation in real and simulated datasets with non-additive interactions. The results showed that the permutation feature importance metric provides more precise feature importance rank estimation in the presence of non-additive interactions, particularly in simulated datasets.
Article
Engineering, Marine
Hong Yang, Xiaodie Yang, Guohui Li
Summary: A dual feature extraction and recognition system is proposed for accurately recognizing ship-radiated noise. The system utilizes pre-processing, dual feature extraction, and recognition techniques. Experimental results show that the system has good classification performance, achieving a recognition rate of up to 97.5%. The system is also successfully applied to simulated chaotic signals and measured aquatic life signals.
Article
Management
Lucas M. Pavelski, Myriam Delgado, Marie-Eleonore Kessaci, Alex A. Freitas
Summary: This paper investigates the use of meta-learning to recommend different stochastic local searches and their parameters to solve permutation flowshop problems. The proposed approach builds a performance database and trains multiple recommendation models to achieve good performance and quality of recommendations. Experiments show that this method outperforms the state-of-the-art algorithm with tuned configuration.
INTERNATIONAL TRANSACTIONS IN OPERATIONAL RESEARCH
(2023)
Review
Cell Biology
Anne M. Bronikowski, Richard P. Meisel, Peggy R. Biga, James R. Walters, Judith E. Mank, Erica Larschan, Gerald S. Wilkinson, Nicole Valenzuela, Ashley Mae Conard, Joao Pedro de Magalhaes, Jingyue (Ellie) Duan, Amy E. Elias, Tony Gamble, Rita M. Graze, Kristin E. Gribble, Jill A. Kreiling, Nicole C. Riddle
Summary: Sex differences in aging, including lifespan, age-associated decline, and physiological markers, vary greatly across animal species. The underlying causes of these differences remain mostly unknown, highlighting the need for further research on the role of sex-determination mechanisms and their impact on aging.
Review
Food Science & Technology
Riccardo Secci, Alexander Hartmann, Michael Walter, Hans Joergen Grabe, Sandra van der Auwera-Palitschka, Axel Kowald, Daniel Palmer, Gerald Rimbach, Georg Fuellen, Israel Barrantes
Summary: This review focuses on the role of diet in protecting against aging, discussing biomarkers based on omics and clinical data, as well as the effects of berry-based interventions. It also emphasizes the importance of individuals' dietary history in geroprotection.
CRITICAL REVIEWS IN FOOD SCIENCE AND NUTRITION
(2023)
Editorial Material
Biotechnology & Applied Microbiology
Joao Pedro de Magalhaes, Alejandro Ocampo
Summary: Cells can be rejuvenated and biological clocks reset through cellular reprogramming, and many companies are now developing therapies using this technique to rejuvenate human beings. However, the current research in rejuvenation is mostly based on in vitro studies, and it remains to be seen if it can be translated into clinical applications.
TRENDS IN BIOTECHNOLOGY
(2022)
Review
Cell Biology
Georg Fuellen, Uwe Walter, Larissa Henze, Jan Bohmert, Daniel Palmer, Soyoung Lee, Clemens A. Schmitt, Henrik Rudolf, Axel Kowald
Summary: The most important predictors for outcomes after ischemic stroke are chronological age and stroke severity, with gender, genetics, and lifestyle/environmental factors also playing a role. Recurrent stroke can be prevented through various therapies and treatment of risk factors. Protein biomarkers, particularly those related to immune-inflammatory, coagulation, and adhesion processes, can provide insight into predicting health deterioration following stroke. These processes also connect stroke to cancer and other conditions, indicating potential overlap in biomarkers.
CELLULAR AND MOLECULAR NEUROBIOLOGY
(2023)
Article
Biochemistry & Molecular Biology
Bowen Song, Xuan Wang, Zhanmin Liang, Jiongming Ma, Daiyun Huang, Yue Wang, Joao Pedro de Magalhaes, Daniel J. Rigden, Jia Meng, Gang Liu, Kunqi Chen, Zhen Wei
Summary: Recent advances in epitranscriptomics have revealed the functional associations between RNA modifications (RMs) and multiple human diseases. This study presents an updated database, RMDisease v2.0, which identifies RM-associated genetic variants that may affect different types of RNA modifications in various organisms. These variants, including disease- and trait-associated genetic variants, may function through perturbations of epitranscriptomic markers. RMDisease v2.0 serves as a valuable resource for studying the genetic drivers of phenotypes in the epitranscriptome layer circuitry.
NUCLEIC ACIDS RESEARCH
(2023)
Article
Oncology
Sina Sender, Ahmad Wael Sultan, Daniel Palmer, Dirk Koczan, Anett Sekora, Julia Beck, Ekkehard Schuetz, Leila Taher, Bertram Brenig, Georg Fuellen, Christian Junghanss, Hugo Murua Escobar
Summary: This study evaluated the impact of simultaneous inhibition of SYK and BET, two key regulators involved in B-cell lymphoma progression. The results showed enhanced anti-proliferative effects and a distinct gene expression profile when SYK and BET were inhibited simultaneously. This suggests that simultaneous inhibition of SYK and BET could be a promising combination therapy for B-cell lymphoma.
Article
Cell Biology
Dominic F. Bennett, Anita Goyala, Cyril Statzer, Charles W. Beckett, Alexander Tyshkovskiy, Vadim N. Gladyshev, Collin Y. Ewald, Joao Pedro de Magalhaes
Summary: Searching for drugs with similar gene expression patterns, rilmenidine was found to extend the lifespan of nematodes and rats, mediated by the I1-imidazoline receptor. This study suggests the potential of rilmenidine as a longevity-promoting drug.
Article
Biochemistry & Molecular Biology
Priyanka Raina, Rodrigo Guinea, Kasit Chatsirisupachai, Ines Lopes, Zoya Farooq, Cristina Guinea, Csaba-Attila Solyom, Joao Pedro de Magalhaes
Summary: Gene co-expression analysis is a powerful method for understanding gene function and regulation. The GeneFriends database provides researchers with updated gene co-expression networks for humans, mice, and other model organisms, based on RNA-seq data. This valuable tool can help researchers decipher the complexity of genomes and assign functions to poorly annotated genes.
NUCLEIC ACIDS RESEARCH
(2023)
Article
Biochemistry & Molecular Biology
Yuxin Zhang, Jie Jiang, Jiongming Ma, Zhen Wei, Yue Wang, Bowen Song, Jia Meng, Guifang Jia, Joao Pedro de Magalhaes, Daniel J. Rigden, Daiyun Hang, Kunqi Chen
Summary: Mapping RNA modifications with advanced technologies has revolutionized our understanding of them. Current modification profiling methods are limited and require selective treatments. Direct RNA sequencing enables the direct study of modifications and has the potential to overcome the limitations of previous methods. The DirectRMDB database provides a fresh perspective on RNA modifications and allows exploration of the epitranscriptome in an isoform-specific manner.
NUCLEIC ACIDS RESEARCH
(2023)
Article
Biochemical Research Methods
Marcelo Rodrigues de Holanda Maia, Alexandre Plastino, Alex Freitas, Joao Pedro de Magalhaes
Summary: This paper proposes two new approaches for dealing with data uncertainty. One approach is to select training instances for each model in an ensemble, and the other is to sample features when splitting a node in a Random Forest training. These approaches are applied to classify ageing-related genes and predict drugs' side effects, and the results show that ensembles based on these approaches achieve better predictive performance.
IEEE-ACM TRANSACTIONS ON COMPUTATIONAL BIOLOGY AND BIOINFORMATICS
(2023)
Article
Biochemistry & Molecular Biology
Xuan Wang, Yuxin Zhang, Kunqi Chen, Zhanmin Liang, Jiongming Ma, Rong Xia, Joao Pedro de Magalhaes, Daniel J. Rigden, Jia Meng, Bowen Song
Summary: With the advancement in mapping N7-methylguanosine (m(7)G) RNA methylation sites, a comprehensive resource (m7Ghub v.2.0) has been developed to study m(7)G modification under various physiological contexts. The resource includes the m7GDB database collecting hundreds of thousands of putative m(7)G sites identified in 23 species, the m7GDiseaseDB hosting m(7)G-associated variants including disease-relevant m(7)G-SNPs, and two enhanced analysis modules for interactive analyses.
NUCLEIC ACIDS RESEARCH
(2023)
Article
Biochemistry & Molecular Biology
Joao Pedro de Magalhaes, Zoya Abidi, Gabriel Arantes dos Santos, Roberto A. Avelar, Diogo Barardo, Kasit Chatsirisupachai, Peter Clark, Evandro A. De-Souza, Emily J. Johnson, Ines Lopes, Guy Novoa, Ludovic Senez, Angelo Talay, Daniel Thornton, Paul Ka Po To
Summary: This article introduces the key features and recent enhancements of the Human Ageing Genomic Resources (HAGR), focusing on its six main databases. These databases cover information related to genes and ageing, longevity, life-history, cellular senescence, and genetic variants associated with human longevity. HAGR also provides various tools and gene expression signatures.
NUCLEIC ACIDS RESEARCH
(2023)
Article
Cell Biology
Cyril Lagger, Eugen Ursu, Anais Equey, Roberto A. Avelar, Angela Oliveira Pisco, Robi Tacutu, Joao Pedro de Magalhaes
Summary: Dysregulation of intercellular communication is a hallmark of aging. Here the authors provide a bioinformatics tool to infer changes in cell-cell signaling and an atlas of age-related communication changes in 23 mouse tissues.