Vancomycin trough concentration in adult patients with periprosthetic joint infection: A machine learning-based covariate model.

AIMS: Although there are various model-based approaches to individualized vancomycin (VCM) administration, few have been reported for adult patients with periprosthetic joint infection (PJI). This work attempted to develop a machine learning (ML)-based model for predicting VCM trough concentration in adult PJI patients. METHODS: The dataset of 287 VCM trough concentrations from 130 adult PJI patients was split into a training set (229) and a testing set (58) at a ratio of 8:2, and an independent external 32 concentrations were collected as a validation set. A total of 13 covariates and the target variable (VCM trough concentration) were included in the dataset. A covariate model was respectively constructed by support vector regression, random forest regression and gradient boosted regression trees and interpreted by SHapley Additive exPlanation (SHAP). RESULTS: The SHAP plots visualized the weight of the covariates in the models, with estimated glomerular filtration rate and VCM daily dose as the 2 most important factors, which were adopted for the model construction. Random forest regression was the optimal ML algorithm with a relative accuracy of 82.8% and absolute accuracy of 67.2% (R2 =.61, mean absolute error = 2.4, mean square error = 10.1), and its prediction performance was verified in the validation set. CONCLUSION: The proposed ML-based model can satisfactorily predict the VCM trough concentration in adult PJI patients. Its construction can be facilitated with only 2 clinical parameters (estimated glomerular filtration rate and VCM daily dose), and prediction accuracy can be rationalized by SHAP values, which highlights a profound practical value for clinical dosing guidance and timely treatment.

Precise Differentiation of Wobble-Type Allele via Ratiometric Design of a Ligase Chain Reaction-Based Electrochemical Biosensor for CYP2C19*2 Genotyping of Clinical Samples.

Due to the comparable stability between the perfect-base pair and the wobble-base pair, a precise differentiation of the wobble-type allele has remained a challenge, often leading to false results. Herein, we proposed a ligase chain reaction (LCR)-based ratiometric electrochemical DNA sensor, namely, R-eLCR, for a precise typing of the wobble-type allele, in which the traditionally recognized "negative" signal of wobble-base pair-mediated amplification was fully utilized as a "positive" one and a ratiometric readout mode was employed to ameliorated the underlying potential external influence and improved its detection accuracy in the typing of the wobble-type allele. The results showed that the ratio between current of methylene blue (IMB) and current of ferrocene (IFc) was partitioned in three regions and three types of wobble-type allele were thus precisely differentiated (AA homozygote: IMB/IFc > 2; GG homozygote: IMB/IFc < 1; GA heterozygote: 1 < IMB/IFc < 2); the proposed R-eLCR successfully discriminated the three types of CYP2C19*2 allele in nine cases of human whole blood samples, which was consistent with those of the sequencing method. These results evidence that the proposed R-eLCR can serve as an accurate and robust alternative for the identification of wobble-type allele, which lays a solid foundation and holds great potential for precision medicine.

Nucleic Acid Amplification by Template-Dominated Click Chemistry for Ultrasensitive DNA/RNA Detection on an Electrochemical Readout Platform.

As an enzyme-free exponential nucleic acid amplification method, the click chemistry-mediated ligation chain reaction (ccLCR) has shown great prospects in the molecular diagnosis. However, the current optics-based ccLCR is challenged by remarkable nonspecific amplification, severely hindering its future application. This study demonstrated that the severe nonspecific amplification was generated probably due to high random collision in the high DNA probe concentration (µM level). To solve this hurdle, a nucleic acid template-dominated ccLCR was constructed using nM-level DNA probes and read on an electrochemical platform (cc-eLCR). Under the optimal conditions, the proposed cc-eLCR detected a low-level nucleic acid target (1 fM) with a single-base resolution. Furthermore, this assay was applied to detect the target of interest in cell extracts with a satisfactory result. The proposed cc-eLCR offers huge possibility for click chemistry-mediated enzyme-free exponential nucleic acid amplification in the application of medical diagnosis and biomedical research.

Cell viability and drug evaluation biosensing system based on disposable AuNPs/MWCNT nanocomposite modified screen-printed electrode for exocytosis dopamine detection.

Cell viability, as an important index to evaluate drug effects, usually was measured by tetrazolium colorimetric assay, playing a key role in drug development and drug screening. Tedious operating procedures, unsatisfactory sensitivity and abominable environments perplex researchers to acquire more detailed in vivo-relevant biological information. Herein, a simple and low-cost cell viability and drug evaluation biosensing system-based on multiwalled carbon nanotubes, gold nanoparticles and Nafion modified screen-printed electrode (SPE) biosensor was constructed for detection of dopamine (DA) released from living cells to evaluate cytotoxicity of antineoplastic drugs such as cisplatin and resveratrol. The biosensing system was demonstrated to display exceptional selectivity, excellent flexibility and good stability toward DA measurement in complex bio-samples. Additionally, the satisfactory recoveries of DA in real samples revealed the reliability and accuracy of the biosensing system in practical application. The IC50 curves respectively obtained by the biosensing system and tetrazolium colorimetric assay provided similar IC50 value but distinctly different dose-effect relationship, which confirmed the enormous potential of the biosensor in cell viability and described drug efficacy profiles in cell function. In short, the cell viability and drug evaluation system using SPE biosensor paves a new way in drug screening and pharmaceutical application to measure bioactive molecule such as DA.

Multimodal Ochratoxin A-Aptasensor Using 3'-FAM-Enhanced Exonuclease I Tool and Magnetic Microbead Carrier.

Thanks to its preparatory ease, close affinity, and low cost, the aptasensor can serve as a promising substitute for antibody-dependent biosensors. However, the available aptasensors are mostly subject to a single-mode readout and the interference of unbound aptamers in solution and non-target-induced transition events. Herein, we proposed a multimodal aptasensor for multimode detection of ochratoxin A (OTA) with cross-validation using the 3'-6-carboxyfluorescein (FAM)-enhanced exonuclease I (Exo I) tool and magnetic microbead carrier. Specifically, the 3'-FAM-labeled aptamer/biotinylated-cDNA hybrids were immobilized onto streptavidin-magnetic microbeads via streptavidin-biotin interaction. With the presence of OTA, an antiparallel G-quadruplex conformation was formed, protecting the 3'-FAM labels from Exo I digestion, and then anti-FAM-horseradish peroxidase (HRP) was bound via specific antigen-antibody affinity; for the aptamers without the protection of OTA, the distal ssDNA was hydrolyzed from 3' → 5', releasing 3'-FAM labels to the solution. Therefore, the OTA was detected by analyzing the "signal-off" fluorescence of the supernatant and two "signal-on" signals in electrochemistry and colorimetry through the detection of the coating magnetic microbeads in HRP's substrate. The results showed that the 3'-FAM labels increased the activity of Exo I, producing a low background due to a more thorough digestion of unbound aptamers. The proposed multimodal aptasensor successfully detected the OTA in actual samples. This work first provides a novel strategy for the development of aptasensors with Exo I and 3'-FAM labels, broadening the application of aptamer in the multimode detection of small molecules.

DNA Nanosieve-Based Regenerative Electrochemical Biosensor Utilizing Nucleic Acid Flexibility for Accurate Allele Typing in Clinical Samples.

Herein, an interface-based DNA nanosieve that has the ability to differentiate ssDNA from dsDNA has been demonstrated for the first time. The DNA nanosieve could be readily built through thiol-DNA's self-assembly on the gold electrode surface, and its cavity size was tunable by varying the concentration of thiol-DNAs. Electrochemical chronocoulometry using [Ru(NH3)6]3+ as redox revealed that the average probe-to-probe separation in the 1 µM thiol-DNA-modified gold electrode was 10.6 ± 0.3 nm so that the rigid dsDNA with a length of ∼17 nm could not permeate the nanosieve, whereas the randomly coiled ssDNA could enter it due to its high flexibility, which has been demonstrated by square wave voltammetry and methylene blue labels through an upside-down hybridization format. After combining the transiently binding characteristic of a short DNA duplex and introducing a regenerative probe (the counterpart of ssDNA), a highly reproducible nanosieve-based E-DNA model was obtained with a relative standard deviation (RSD) as low as 2.7% over seven cycles. Finally, we built a regenerative nanosieve-based E-DNA sensor using a ligation cycle reaction as an ssDNA amplification strategy and realized one-sensor-based continuous measurement to multiple clinical samples with excellent allele-typing performance. This work holds great potential in low-cost and high-throughput analysis between biosensors and biochips and also opens up a new avenue in nucleic acid flexibility-based DNA materials for future applications in DNA origami and molecular logic gates.

Ultrasensitive Detection of RNA with Single-Base Resolution by Coupling Electrochemical Sensing Strategy with Chimeric DNA Probe-Aided Ligase Chain Reaction.

Accurate and sensitive detection of single-base mutations in RNAs is of great value in basic studies of life science and medical diagnostics. However, the current available RNA detection methods are challenged by heterogeneous clinical samples in which trace RNA mutants usually existed in a large pool of normal wild sequences. Thus, there is still great need for developing the highly sensitive and highly specific methods in detecting single-base mutations of RNAs in heterogeneous clinical samples. In the present study, a new chimeric DNA probe-aided ligase chain reaction-based electrochemical method (cmDNA-eLCR) was developed for RNA mutation detection through the BSA-based carrier platform and the horseradish peroxidase-hydrogen peroxide-tetramethylbenzidine (HRP-H2O2-TMB) system. The denaturing polyacrylamide gel electrophoresis and a fluorophore-labeled probe was ingeniously designed to demonstrate the advantage of cmDNA in ligation to normal DNA templated by RNA with the catalysis of T4 RNA ligase 2 as well as its higher selectivity than DNA ligase system. Finally, the proposed cmDNA-eLCR, compared with the traditional eLCR, showed excellent performance in discriminating single base-mismatched sequences, where the signal response for mismatched targets at a high concentration could overlap completely with that for the blank control. Besides, this cmDNA-eLCR assay had a wide linear range crossing six orders of magnitude from 1.0 × 10-15 to1.0 × 10-10 M with a limit of detection as low as 0.6 fM. Furthermore, this assay was applied to detect RNA in real sample with a satisfactory result, thereby demonstrating its great potential in diagnosis of RNA-related diseases.

A novel ligase chain reaction-based electrochemical biosensing strategy for highly sensitive point mutation detection from human whole blood.

Challenged by the detection of trace amounts of mutants and disturbance from endogenous substances in clinical samples, herein, we present a novel electrochemical biosensor based on ligase chain reaction (eLCR) via the thermostable ligase with high mutation recognizing ability. The lengthened double-stranded DNAs exponentially generated via LCR were uniformly distributed on a bovine serum albumin-modified gold electrode, in which the phosphate buffer was tactfully added to remove adsorbed uninterested-probes, and thereafter the amperometry current was collected for the specific binding of streptavidin-poly-HRP and subsequent catalysis in the 3, 3', 5, 5'-tetramethylbenzidine substrate that contained hydrogen peroxide. It found that, under optimized conditions, the proposed biosensor exhibited a high selectivity of mutant targets from the 104-fold excess of co-existent wild targets within a detection limit of 0.5 fM. Impressively, without the involvement of pre-PCR, the homozygous mutants were specifically distinguished from the wild genotype of CYP2C19*2 allele in human whole blood samples. Therefore, the proposed eLCR, due to its advantages in simple primer design, operational ease and ease of miniaturization, has demonstrated its considerable potential for point-of-care testing in the diagnosis of point mutation-related diseases and personalized medicine.

Ultrasensitive Electrochemical Biosensor Developed by Probe Lengthening for Detection of Genomic DNA in Human Serum.

As an alternative to most of the reported nucleic acid amplification-based electrochemical DNA biosensors used for detection of trace levels of genomic DNA, we herein present a novel detection concept. The proposed system involves the conversion of two short double-stranded DNAs (dsDNAs), labeled with a thiol-tag or biotin-tag, into a single integrated dsDNA containing thiol and biotin at both terminals in the presence of target DNA through ligase chain reaction (LCR) and followed by the immobilization of these integrated dsDNAs on a bovine serum albumin (BSA)-modified gold electrode surface. Owing to rapid depletion of the two short dsDNAs via LCR, the integrated dsDNAs were generated in an exponential manner so that this sensoring approach offered a limit of detection of 25 yoctomoles (15 copies in 50 µL sample volumes), a high discrimination of single-base mismatch and a wide linear concentration range (across 6 orders of magnitude) for target DNA. Significantly, the proposed sensor, which has simplicity in operation and ease of miniaturization, detected the target of interest in total nucleic acid extracts derived from clinical serum samples with excellent results, thereby demonstrating its considerable diagnostic potential in fields ranging from virus detection to the diagnosis of genetic diseases.