Small molecule metabolites: discovery of biomarkers and therapeutic targets

Signal Transduction and Targeted Therapy, Journal Year: 2023, Volume and Issue: 8(1)

Published: March 20, 2023

Metabolic abnormalities lead to the dysfunction of metabolic pathways and metabolite accumulation or deficiency which is well-recognized hallmarks diseases. Metabolite signatures that have close proximity subject's phenotypic informative dimension, are useful for predicting diagnosis prognosis diseases as well monitoring treatments. The lack early biomarkers could poor serious outcomes. Therefore, noninvasive methods with high specificity selectivity desperately needed. Small molecule metabolites-based metabolomics has become a specialized tool biomarker pathway analysis, revealing possible mechanisms human various deciphering therapeutic potentials. It help identify functional related variation delineate biochemical changes indicators pathological damage prior disease development. Recently, scientists established large number profiles reveal underlying networks target exploration in biomedicine. This review summarized analysis on potential value small-molecule candidate metabolites clinical events, may better diagnosis, prognosis, drug screening treatment. We also discuss challenges need be addressed fuel next wave breakthroughs.

Language: Английский

---

Harnessing Machine Learning for Quantifying Vesicoureteral Reflux: A Promising Approach for Objective Assessment

International Journal of Online and Biomedical Engineering (iJOE), Journal Year: 2024, Volume and Issue: 20(11), P. 123 - 145

Published: Aug. 8, 2024

In this study, we evaluated the performance of various machine-learning models on multiple datasets labeled GR1, GR2, GR3, GR4, and GR5. We assessed using a range evaluation metrics, including AUC, CA, F1, precision, recall, MCC, specificity, log loss. The examined were logistic regression, decision tree, kNN, random forest, gradient boosting, neural network, AdaBoost, stochastic descent. results indicate that all consistently demonstrated outstanding across datasets, with most achieving perfect scores in metrics. exhibited high accuracy effectiveness accurately classifying instances. Although forests displayed slightly lower some theyi still maintained an overall level accuracy. findings highlight models’ ability to effectively learn underlying patterns within data make accurate predictions. low loss values further confirmed precise estimation probabilities. Consequently, these possess strong potential for practical applications domains, offering reliable robust classification capabilities.

Language: Английский

---

Investigation of a targeted panel of gut microbiome–derived toxins in children with chronic kidney disease

Pediatric Nephrology, Journal Year: 2025, Volume and Issue: unknown

Published: Jan. 16, 2025

Language: Английский

---

Interpretable machine learning for predicting chronic kidney disease progression risk

Digital Health, Journal Year: 2024, Volume and Issue: 10

Published: Jan. 1, 2024

Objective Chronic kidney disease (CKD) poses a major global health burden. Early CKD risk prediction enables timely interventions, but conventional models have limited accuracy. Machine learning (ML) enhances prediction, interpretability is needed to support clinical usage with both in diagnostic and decision-making. Methods A cohort of 491 patients data was collected for this study. The dataset randomly split into an 80% training set 20% testing set. To achieve the first objective, we developed four ML algorithms (logistic regression, random forests, neural networks, eXtreme Gradient Boosting (XGBoost)) classify two classes—those who progressed stages 3–5 during follow-up (positive class) those did not (negative class). For classification task, area under receiver operating characteristic curve (AUC-ROC) used evaluate model performance discriminating between classes. survival analysis, Cox proportional hazards regression (COX) forests (RSFs) were employed predict progression, concordance index (C-index) integrated Brier score evaluation. Furthermore, variable importance, partial dependence plots, restrict cubic splines interpret models’ results. Results XGBOOST demonstrated best predictive progression AUC-ROC 0.867 (95% confidence interval (CI): 0.728–0.100), outperforming other algorithms. In RSF showed slightly better discrimination calibration on test compared COX, indicating generalization new data. Variable importance analysis identified estimated glomerular filtration rate, age, creatinine as most important predictors analysis. Further revealed non-linear associations age suggesting higher risks aged 52–55 65–66 years. association cholesterol levels also non-linear, lower observed when range 5.8–6.4 mmol/L. Conclusions Our study effectiveness interpretable predicting progression. comparison COX highlighted advantages particularly handling non-linearity high-dimensional By leveraging unraveling factor relationships, contrasting techniques, exposing associations, significantly advances enable enhanced

Language: Английский

---

Artificial intelligence: a new field of knowledge for nephrologists?

Clinical Kidney Journal, Journal Year: 2023, Volume and Issue: 16(12), P. 2314 - 2326

Published: July 29, 2023

Artificial intelligence (AI) is a science that involves creating machines can imitate human and learn. AI ubiquitous in our daily lives, from search engines like Google to home assistants Alexa and, more recently, OpenAI with its chatbot. improve clinical care research, but use requires solid understanding of fundamentals, the promises perils algorithmic fairness, barriers solutions implementation, pathways developing an AI-competent workforce. The potential field nephrology vast, particularly areas diagnosis, treatment prediction. One most significant advantages ability diagnostic accuracy. Machine learning algorithms be trained recognize patterns patient data, including lab results, imaging medical history, order identify early signs kidney disease thereby allow timely diagnoses prompt initiation plans outcomes for patients. In short, holds promise advancing personalized medicine new levels. While has tremendous potential, there are also challenges data access quality, privacy security, bias, trustworthiness, computing power, integration legal issues. European Commission's proposed regulatory framework technology will play role ensuring safe ethical implementation these technologies healthcare industry. Training nephrologists fundamentals imperative because traditionally, decision-making pertaining prognosis renal patients relied on ingrained practices, whereas serves as powerful tool swiftly confidently synthesizing this information.

Language: Английский

---

Metabolites Associated With Uremic Symptoms in Patients With CKD: Findings From the Chronic Renal Insufficiency Cohort (CRIC) Study

American Journal of Kidney Diseases, Journal Year: 2024, Volume and Issue: 84(1), P. 49 - 61.e1

Published: Jan. 23, 2024

Language: Английский

---

Associations of neighborhood sociodemographic environment with mortality and circulating metabolites among low-income black and white adults living in the southeastern United States

BMC Medicine, Journal Year: 2024, Volume and Issue: 22(1)

Published: June 18, 2024

Residing in a disadvantaged neighborhood has been linked to increased mortality. However, the impact of residential segregation and social vulnerability on cause-specific mortality is understudied. Additionally, circulating metabolic correlates sociodemographic environment remain unexplored. Therefore, we examined multiple metrics, i.e., deprivation index (NDI), (RSI), (SVI), with all-cause cardiovascular disease (CVD) cancer-specific metabolites Southern Community Cohort Study (SCCS).

Language: Английский

---

Longitudinal Plasma Metabolome Patterns and Relation to Kidney Function and Proteinuria in Pediatric CKD

Clinical Journal of the American Society of Nephrology, Journal Year: 2024, Volume and Issue: 19(7), P. 837 - 850

Published: May 6, 2024

Key Points Longitudinal untargeted metabolomics. Children with CKD have a circulating metabolome that changes over time. Background Understanding plasma patterns in relation to changing kidney function pediatric is important for continued research identifying novel biomarkers, characterizing biochemical pathophysiology, and developing targeted interventions. There are limited number of studies longitudinal metabolomics virtually none CKD. Methods The study multi-institutional, prospective cohort enrolled children aged 6 months 16 years eGFR 30–90 ml/min per 1.73 m 2 . Untargeted profiling was performed on samples from the baseline, 2-, 4-year visits. were technologic updates metabolomic platform used between baseline follow-up assays. Statistical approaches adopted avoid direct comparison measurements. To identify metabolite associations or urine protein-creatinine ratio (UPCR) among all three time points, we applied linear mixed-effects (LME) models. metabolites associated time, LME models 2- data. We regression analysis examine change level (∆level) (∆eGFR) UPCR (∆UPCR). reported significance basis both false discovery rate (FDR) <0.05 P < 0.05. Results 1156 person-visits ( N : baseline=626, 2-year=254, 4-year=276) included. 622 standardized measurements at points. In modeling, 406 343 FDR <0.05, respectively. Among 530 person-visits, 158 showed differences <0.05. For participants complete data visits n =123), report 35 ∆level–∆eGFR significant no ∆level–∆UPCR 0.05 modeling Conclusions characterized large population. Many these signals been progression, etiology, proteinuria previous Biomarkers Consortium studies. also detected.

Language: Английский

---

Artificial intelligence in cardiac metabolism: the next frontier in cardiovascular health

Metabolism and Target Organ Damage, Journal Year: 2025, Volume and Issue: 5(1)

Published: Jan. 7, 2025

In this article, we aim to explore the rapidly developing role of artificial intelligence (AI) in cardiac metabolism research, highlighting its impact on biomarker discovery, precision medicine, and patient stratification. Cardiac metabolism, a key determinant cardiovascular health, is often disrupted diseases (CVDs) like heart failure coronary artery disease. AI’s ability process analyze large-scale data offers new chances for understanding addressing these metabolic dysfunctions. By integrating up-to-date technologies with molecular clinical insights, AI enables achievement personalized treatments, more accurate diagnostics, discovery potential novel therapeutic targets. The main challenges include ethical concerns around privacy, algorithmic bias, need representative datasets. Future directions focus transparent, accountable, collaborative models that integrate enable real-time monitoring, ensuring fairness accessibility healthcare. As continues evolve, advancing care expected grow, offering trends research.

Language: Английский

---

Kinetic-pharmacodynamic model to predict post-rituximab B-cell repletion as a predictor of relapse in pediatric idiopathic nephrotic syndrome

Frontiers in Pharmacology, Journal Year: 2025, Volume and Issue: 15

Published: Jan. 7, 2025

Rituximab has proven efficacy in children with idiopathic nephrotic syndrome (INS). However, vast majority of inevitably experience relapse B-cell repletion, necessitating repeat course rituximab, which may increase the risk adverse effects. The timing additional dosing and optional regimen rituximab pediatric patients INS have yet to be determined. This study aimed identify factors that influence disease repletion provide tailored treatment. LASSO random survival forest were performed on 143 screen covariates then included Cox regression model determine biomarkers establish a nomogram. A kinetic-pharmacodynamic (K-PD) was developed 59 characterize time CD19+ after Monte Carlo simulation conducted explore mini-dose larger intervals. Nomogram contained 7 predictors including neutrophil-to-lymphocyte ratio, duration depletion, disease, urine immunoglobulin G creatinine transferrin, maintenance immunosuppressant hemoglobin. As direct PD indicator, each 1-month depletion decreased by 21.4% (HR = 0.786; 95% CI: 0.635-0.972; p 0.026). K-PD predicted t1/2 (CV%) 11.6 days (17%) 173.3 (22%), respectively. Immunoglobulin is an important covariate ED50. Simulation intervals (three 150 mg every 2 monthly) indicted longer (>7 months) compared standard regimen. nomogram indicated optimal infusion before provided regimens for reduce safety risks financial burden.

Language: Английский

---