Search

Aims: Few personalized monitoring models for valproic acid (VPA) in pediatric epilepsy patients (PEPs) incorporate machine learning (ML) algorithms. This study aimed to develop an ensemble ML model for VPA monitoring to enhance clinical precision of VPA usage. Methods: A dataset comprising 366 VPA trough concentrations from 252 PEPs, along with 19 covariates and the target variable (VPA trough concentration), was refined by Spearman correlation and multicollinearity testing (366 × 11). The dataset was split into a training set (292) and testing set (74) at a ratio of 8:2. An ensemble model was formulated by Gradient Boosting Regression Trees (GBRT), Random Forest Regression (RFR), and Support Vector Regression (SVR), and assessed by SHapley Additive exPlanations (SHAP) analysis for covariate importance. The model was optimized for R2, relative accuracy, and absolute accuracy, and validated against two independent external datasets (32 in-hospital and 28 out-of-hospital dataset). Results: Using the R2 weight ratio of GBRT, RFR and SVR optimized at 5:2:3, the ensemble model demonstrated superior performance in terms of relative accuracy (87.8%), absolute accuracy (78.4%), and R2 (0.50), while also exhibiting a lower Mean Absolute Error (9.87) and Root Mean Squared Error (12.24), as validated by the external datasets. Platelet count (PLT) and VPA daily dose were identified as pivotal covariates. Conclusion: The proposed ensemble model effectively monitors VPA trough concentrations in PEPs. By integrating covariates across various ML algorithms, it delivers results closely aligned with clinical practice, offering substantial clinical value for the guided use of VPA. [ABSTRACT FROM AUTHOR]

Currently, the reported vancomycin (VCM) aptamers, including the 3- (K d  =  9.13 × 10 -6 m) and 4-truncated variants (K d = 45.5 × 10 -6 m), are engineered via stem truncation of the VCM parent aptamer, which inevitably compromises their affinities, thus affecting their clinical application within the VCM therapeutic window of 6.9-13.8 × 10 -6 m. Herein, the binding pocket of the VCM parent aptamer is elucidated for the first time and we implemented the Post-SELEX modification strategy involving truncation and mutagenesis to refined the VCM parent aptamer. This yielded a VCM aptamer (ABC20-11) with an intramolecular G-triplex, an enhanced thioflavin T (ThT) fluorescence intensity, and an improved affinity (K d = 0.591 × 10 -6 m) and specificity (one-methyl level) for VCM. Utilizing a portable fluorescence detector specifically designed for rapidly detecting VCM concentration and leveraging the competitive binding between VCM and ThT to ABC20-11, a label-free fluorescent aptasensor is developed. This aptasensor exhibits exceptional analytical performances across various clinical samples (serum, cerebrospinal fluid, and joint fluid), with corresponding linear ranges of 0.5-50, 0.5-40, and 0.5-50 × 10 -6 m and detection limits at 0.11, 0.12, and 0.16 × 10 -6 m, respectively. Consequently, the proposed VCM aptasensor displays considerable clinical value and potential for use in rapid VCM detection., (© 2024 Wiley‐VCH GmbH.)

Aims: Few personalized monitoring models for valproic acid (VPA) in pediatric epilepsy patients (PEPs) incorporate machine learning (ML) algorithms. This study aimed to develop an ensemble ML model for VPA monitoring to enhance clinical precision of VPA usage., Methods: A dataset comprising 366 VPA trough concentrations from 252 PEPs, along with 19 covariates and the target variable (VPA trough concentration), was refined by Spearman correlation and multicollinearity testing (366 × 11). The dataset was split into a training set (292) and testing set (74) at a ratio of 8:2. An ensemble model was formulated by Gradient Boosting Regression Trees (GBRT), Random Forest Regression (RFR), and Support Vector Regression (SVR), and assessed by SHapley Additive exPlanations (SHAP) analysis for covariate importance. The model was optimized for R 2 , relative accuracy, and absolute accuracy, and validated against two independent external datasets (32 in-hospital and 28 out-of-hospital dataset)., Results: Using the R 2 weight ratio of GBRT, RFR and SVR optimized at 5:2:3, the ensemble model demonstrated superior performance in terms of relative accuracy (87.8%), absolute accuracy (78.4%), and R 2 (0.50), while also exhibiting a lower Mean Absolute Error (9.87) and Root Mean Squared Error (12.24), as validated by the external datasets. Platelet count (PLT) and VPA daily dose were identified as pivotal covariates., Conclusion: The proposed ensemble model effectively monitors VPA trough concentrations in PEPs. By integrating covariates across various ML algorithms, it delivers results closely aligned with clinical practice, offering substantial clinical value for the guided use of VPA., Competing Interests: The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest., (Copyright © 2024 Chen, Lin, Chen, Zhang, Wu, Huang, Rao, Lu and Liu.)