163559 research outputs found
Sort by
Multi-label classification with XGBoost for metabolic pathway prediction
Get PDFBackground
Metabolic pathway prediction is one possible approach to address the problem in system biology of reconstructing an organisms metabolic network from its genome sequence. Recently there have been developments in machine learning-based pathway prediction methods that conclude that machine learning-based approaches are similar in performance to the most used method, PathoLogic which is a rule-based method. One issue is that previous studies evaluated PathoLogic without taxonomic pruning which decreases its performance.
Results
In this study, we update the evaluation results from previous studies to demonstrate that PathoLogic with taxonomic pruning outperforms previous machine learning-based approaches and that further improvements in performance need to be made for them to be competitive. Furthermore, we introduce mlXGPR, a XGBoost-based metabolic pathway prediction method based on the multi-label classification pathway prediction framework introduced from mlLGPR. We also improve on this multi-label framework by utilizing correlations between labels using classifier chains. We propose a ranking method that determines the order of the chain so that lower performing classifiers are placed later in the chain to utilize the correlations between labels more. We evaluate mlXGPR with and without classifier chains on single-organism and multi-organism benchmarks. Our results indicate that mlXGPR outperform other previous pathway prediction methods including PathoLogic with taxonomic pruning in terms of hamming loss, precision and F1 score on single organism benchmarks.
Conclusions
The results from our study indicate that the performance of machine learning-based pathway prediction methods can be substantially improved and can even outperform PathoLogic with taxonomic pruning.This work was supported by the National Research Foundation of Korea(NRF) grant funded by the Korea government(Ministry of Science and ICT) (No. RS-2023-00268071
Mapping cancer biology in space: applications and perspectives on spatial omics for oncology
Get PDFAbstract
Technologies to decipher cellular biology, such as bulk sequencing technologies and single-cell sequencing technologies, have greatly assisted novel findings in tumor biology. Recent findings in tumor biology suggest that tumors construct architectures that influence the underlying cancerous mechanisms. Increasing research has reported novel techniques to map the tissue in a spatial context or targeted sampling-based characterization and has introduced such technologies to solve oncology regarding tumor heterogeneity, tumor microenvironment, and spatially located biomarkers. In this study, we address spatial technologies that can delineate the omics profile in a spatial context, novel findings discovered via spatial technologies in oncology, and suggest perspectives regarding therapeutic approaches and further technological developments
Unsupervised generation of fashion editorials using deep generative model
No full textThis research intended to establish a new fashion-related artificial intelligence research topic concerning fashion editorials which could induce streams of further studies. A new fashion editorial dataset, which is a prerequisite in training an AI model, has been established in this study to meet the research purpose. A total of over 150K fashion editorials were initially collected and processed to satisfy necessary dataset conditions. A novel dataset of fashion editorials consisting of approximately 60K editorials is proposed through the process. In order to prove the adequacy of the new dataset, data distribution was analyzed and a generative model was selected and trained to attest that new fashion editorials can be created with the proposed editorial dataset. The results generated by the trained model were qualitatively investigated. The model has shown to have learned various features that compose editorials with the dataset, successfully generating fashion editorials. Quantitative evaluation with FID scores was conducted to support the selection of the generative model used for the qualitative assessment
Correction: Glial cell proteome using targeted quantitative methods for potential multi-diagnostic biomarkers
No full textWearable and Implantable Light-Emitting Diodes and Their Biomedical Applications
No full textLight-mediated healthcare, including monitoring and therapy techniques, has evolved significantly over the past decades, owing to their advantages in minimal invasiveness and relatively low side effects. However, most of the clinically available light sources have suffered from their bulky size and rigidity, making them impractical for continuous on-body-type health monitoring and treatment applications. Recently, wearable and implantable healthcare systems using deformable light emitting diodes (LEDs) have been developed, which can be seamlessly integrated to any part of the body and exhibit the potential to improve efficacy of personalized and mobile photo-medicine. In this review, we discuss recent advances in wearable and implantable LED technologies for various biomedical applications. First, we present an overview of the stateof-the-art high-performance wearable and implantable LEDs, especially focusing on their light-emitting materials and unique device design approaches. We also discuss strategies for integrated light-based healthcare systems. Then, we discuss prominent examples of light-medicated sensing and therapeutic applications, whose accuracy and efficacy can be enhanced by wearable and implantable LEDs, ranging from non-invasive oxygen level sensing to various phototherapies. Finally, we conclude this review with a brief outlook on the future technologies.N
Landslide susceptibility assessment of South Korea using stacking ensemble machine learning
No full textBackground
Landslide susceptibility assessment (LSA) is a crucial indicator of landslide hazards, and its accuracy is improving with the development of artificial intelligence (AI) technology. However, the AI algorithms are inconsistent across regions and strongly dependent on input variables. Additionally, LSA must include historical data, which often restricts the assessment to the local scale and single landslide events.
Methods
In this study, we performed an LSA for the entirety of South Korea. A total of 30 input variables were constructed, consisting of 9 variables from past climate model data MK-PRISM, 12 topographical factors, and 9 environmental factors. Sixteen machine learning algorithms were used as basic classifiers, and a stacking ensemble was used on the four algorithms with the highest area under the curve (AUC). Additionally, a separate assessment model was established for areas with a risk of landslides affecting areas larger than 1 ha.
Results
The highest-performing classifier was CatBoost, with an AUC of ~ 0.89 for both assessments. Among the input variables, distance of road, daily maximum precipitation, digital elevation model, and soil depth were the most influential. In all landslide events, CatBoost, lightGBM, XGBoost, and Random Forest had the highest AUC in descending order; in large landslide events, the order was CatBoost, XGBoost, Extra Tree, and lightGBM. The stacking ensemble enabled the construction of two landslide susceptibility maps.
Conclusions
Our findings provide a statistical method for constructing a high-resolution (30 m) landslide susceptibility map on a country scale using diverse natural factors, including past climate data.This study was carried out with the support of the R&D Program for Forest Science Technology (Project No. 2021341A00-2323-CD01) provided by the Korea Forest Service (Korea Forestry Promotion Institute
Worrying About Our Children: Parental Worry and Psychological Well-Being Among Korean Middle-Aged Couples
No full textWhen children reach a certain age of maturity, middle-aged parents often reflect on their parenting, harboring continuous worries about their adult children. These parenting experiences are also shared within couples and continue to impact parents' well-being. Utilizing couple data from the 2010 Korean Baby Boomer Panel Study, we examined the dyadic associations of worry about child issues and psychological well-being among middle-aged couples (N = 1,091; aged 47-55) who have at least one adult child (M-age = 23.13 years). Results from the actor-partner interdependence model showed that one's own parental worry was significantly associated with psychological well-being for both husbands and wives (i.e., actor effects). Further, wives' worry about children was significantly associated with husbands' psychological well-being (i.e., partner effects)-but not vice versa. These findings highlight that aspects of parenting not only impact children but also extend to the linked lives of midlife parents themselves. Research on parental experiences at the couple level may inform interventions to enhance middle-aged parents' well-being.N
A novel formula for predicting the ultimate compressive strength of the cylindrically curved plates
No full text"The present study aims to develop an empirical formula to predict the ultimate compressive strength of unstiffened cylindrically curved plates. Drawing from an extensive analysis of 400 unique curved plate scenarios under longitudinal compression, we investigated critical parameters: the flank angle (denoted as θ), plate aspect ratio (denoted as a/b), and plate slenderness ratio (denoted as β). The ANSYS Nonlinear Finite Element Method (NLFEM) was employed to assess each scenario, considering the average level of initial imperfections (denoted as 0.1β2t) and configurations of one-bay and one-span. It is important to note that the models were designed without accounting for the effects of residual stresses. The simulation data generated from this analysis served as the foundation for developing our empirical formula. The proposed formula strongly agreed with the numerical simulations and experimental test results. This research provides structural engineers with a reliable predictive tool, aiding in more accurate predictions of the ultimate limit state (ULS) of curved plates during early design phases.Y
Neoadjuvant and adjuvant pembrolizumab plus chemotherapy in locally advanced gastric or gastro-oesophageal cancer (KEYNOTE-585): an interim analysis of the multicentre, double-blind, randomised phase 3 study
No full textBackground: The benefit of combination neoadjuvant and adjuvant chemotherapy and immune checkpoint inhibition in patients with locally advanced, resectable gastric or gastro-oesophageal adenocarcinoma is unknown. We assess the antitumor activity of neoadjuvant and adjuvant pembrolizumab plus chemotherapy in patients with locally advanced resectable gastric or gastro-oesophageal adenocarcinoma. Methods: The KEYNOTE-585 study is a multicentre, randomised, placebo-controlled, double-blind, phase 3 study done at 143 medical centres in 24 countries. Eligible patients were aged 18 years or older with untreated, locally advanced, resectable gastric or gastro-oesophageal adenocarcinoma, and an Eastern Cooperative Oncology Group performance status 0–1. Patients were randomly assigned (1:1) by an interactive voice response system and integrated web response system to neoadjuvant pembrolizumab 200 mg intravenously or placebo (saline) plus cisplatin-based doublet chemotherapy (main cohort) every 3 weeks for 3 cycles, followed by surgery, adjuvant pembrolizumab or placebo plus chemotherapy for 3 cycles, then adjuvant pembrolizumab or placebo for 11 cycles. A small cohort was also randomly assigned (1:1) to pembrolizumab or placebo plus fluorouracil, docetaxel, and oxaliplatin (FLOT)-based chemotherapy (FLOT cohort) every 2 weeks for four cycles, followed by surgery, adjuvant pembrolizumab, or placebo plus FLOT for four cycles, then adjuvant pembrolizumab or placebo for 11 cycles. Patients were stratified by geographic region, tumour stage, and chemotherapy backbone. Primary endpoints were pathological complete response (reviewed centrally), event-free survival (reviewed by the investigator), and overall survival in the intention-to-treat population, and safety assessed in all patients who received at least one dose of study treatment. The study is registered at ClinicalTrials.gov, NCT03221426, and is closed to accrual. Findings: Between Oct 9, 2017, and Jan 25, 2021, of 1254 patients screened, 804 were randomly assigned to the main cohort, of whom 402 were assigned to the pembrolizumab plus cisplatin-based chemotherapy group and 402 to the placebo plus cisplatin-based chemotherapy group, and 203 to the FLOT cohort, of whom 100 were assigned to the pembrolizumab plus FLOT group and 103 to placebo plus FLOT group. In the main cohort of 804 participants, 575 (72%) were male and 229 (28%) were female. In the main cohort, after median follow-up of 47·7 months (IQR 38·0–54·8), pembrolizumab was superior to placebo for pathological complete response (52 [12·9%; 95% CI 9·8–16·6] of 402 vs eight [2·0%; 0·9–3·9] of 402; difference 10·9%, 95% CI 7·5 to 14·8; p<0·00001). Median event-free survival was longer with pembrolizumab versus placebo (44·4 months, 95% CI 33·0 to not reached vs 25·3 months, 20·6 to 33·9; hazard ratio [HR] 0·81, 95% CI 0·67 to 0·99; p=0·0198) but did not meet the threshold for statistical significance (p=0·0178). Median overall survival was 60·7 months (95% CI 51·5 to not reached) in the pembrolizumab group versus 58·0 months (41·5 to not reached) in the placebo group (HR 0·90, 95% CI 0·73 to 1·12; p=0·174). Grade 3 or worse adverse events of any cause occurred in 312 (78%) of 399 patients in the pembrolizumab group and 297 (74%) of 400 patients in the placebo group; the most common were nausea (240 [60%] vs 247 [62%]), anaemia (168 [42%] vs 158 [40%]), and decreased appetite (163 [41%] vs 172 [43%]). Treatment-related serious adverse events were reported in 102 (26%) and 97 (24%) patients. Treatment-related adverse events that led to death occurred in four (1%) patients in the pembrolizumab group (interstitial ischaemia, pneumonia, decreased appetite, and acute kidney injury [n=1 each]) and two (<1%) patients in the placebo group (neutropenic sepsis and neutropenic colitis [n=1 each]). Interpretation: Although neoadjuvant and adjuvant pembrolizumab versus placebo improved the pathological complete response, it did not translate to significant improvement in event-free survival in patients with untreated, locally advanced resectable gastric or gastro-oesophageal cancer. Funding: Merck Sharp & Dohme.N