SERS and advanced chemometrics – Utilization of Siamese neural network for picomolar identification of beta-lactam antibiotics resistance gene fragment

Antibiotic-resistant bacteria aerosol in a Caribbean coastal city: Pre- and post- COVID-19 lockdown

This study assessed the prevalence and spatial distribution of viable ultrafine and fine antibiotic-resistant bacteria aerosols (ARB) in the Metropolitan Area of Barranquilla, Colombia, pre- and post-lockdown (September 2019 to December 2020). Samples were systematically collected from urban, suburban, and rural sites using a six-stage viable cascade impactor. We employed logistic regression and Bayesian Neural Network Classifiers to analyze meteorological variables' influence on antibiotic resistance persistence. The lockdown led to a significant decrease (76 %) in overall bacterial aerosol concentrations, likely due to reduced human activity. The most significant reduction (82 %) was observed at Peace Square. Bacillus cereus was the most prevalent species, showing high concentrations at all sampling sites. Other species, like Leifsonia aquatica and Staphylococcus lentus, were linked to wastewater effluents and agricultural activities. Despite the overall decrease in bacterial aerosols, antibiotic-resistant bacteria remained high, particularly in highly impacted urban areas like the Barranquilla Riverwalk. Bacillus cereus exhibited resistance to multiple antibiotics, including commonly used ones like Ampicillin and Penicillin G. Resistance to newer antibiotics like Vancomycin was rare. Peace Square, a high-traffic urban area, showed elevated resistance rates in the deeper respiratory regions compared to other locations. Our findings indicate that while overall concentration levels decreased, the threat of antibiotic resistance in bacterial bioaerosols persists, emphasizing the need for continuous monitoring and targeted public health interventions in urban areas.

Methods for Rapid Evaluation of Microbial Antibiotics Resistance

Antibiotic resistance is a major challenge for public health systems worldwide. Rapid and effective identification of bacterial strains is critical for reducing the use of antibiotics and restricting the spread of antibiotic-resistant microorganisms. Various approaches have been developed in recent years for rapid bacterial identification and antibiotic susceptibility testing (AST), such as Raman spectroscopy, single cell image analysis, microfluidic techniques, mass spectrometry analysis, use of high-sensitive luminescent and fluorescent tags, impedance-based detection, and others. This review describes the methods developed for rapid bacterial identification and assessment of their antibiotic susceptibility, including general principles, specific problems, and future prospects.

Surface enhanced Raman spectroscopy and machine learning for identification of beta-lactam antibiotics resistance gene fragment in bacterial plasmid

BACKGROUND

Antibiotic resistance stands as a critical medical concern, notably evident in commonly prescribed beta-lactam antibiotics. The imperative need for expeditious and precise early detection methods underscores their role in facilitating timely intervention, curbing the propagation of antibiotic resistance, and enhancing patient outcomes.

RESULTS

This study introduces the utilization of surface-enhanced Raman spectroscopy (SERS) in tandem with machine learning (ML) for the sensitive detection of characteristic gene fragments responsible for antibiotic resistance appearance and spreading. To make the detection procedure close to the real case, we used bacterial plasmids as starting biological objects, containing or not the characteristic gene fragment (up to 1:10 ratio), encoding beta-lactam antibiotics resistance. The plasmids were subjected to enzymatic digestion and without preliminary purification or isolation the created fragments were captured by functional SERS substrates. Based on subsequent SERS measurements, a database was created for the training and validation of ML. Method validation was performed using separately measured spectra, which did not overlap with the database used for ML training. To check the efficiency of recognising the target fragment, control experiments involved bacterial plasmids containing different resistance genes, the use of inappropriate enzymes, or the absence of plasmid.

SIGNIFICANCE

SERS-ML allowed express detection of bacterial plasmids containing a characteristic gene fragment up to the 10-7 concentration of the initial plasmid, despite the complex composition of the biological sample, including the presence of interfering plasmids. Our approach offers a promising alternative to existing methods for monitoring antibiotic-resistant bacteria, characterized by its simplicity, low detection limit, and the potential for rapid and straightforward analysis.

Label-free SERS-ML detection of cocaine trace in human blood plasma

The widespread consumption of cocaine poses a significant threat to modern society. The most effective way to combat this problem is to control the distribution of cocaine, based on its accurate and sensitive detection. Here, we proposed the detection of cocaine in human blood plasma using a combination of surface enhanced Raman spectroscopy and machine learning (SERS-ML). To demonstrate the efficacy of our proposed approach, cocaine was added into blood plasma at various concentrations and drop-deposited onto a specially prepared disposable SERS substrate. SERS substrates were created by deposition of metal nanoclusters on electrospun polymer nanofibers. Subsequently, SERS spectra were measured and as could be expected, the manual distinguishing of cocaine from the spectra proved unfeasible, as its signal was masked by the background signal from blood plasma molecules. To overcome this issue, a database of SERS spectra of cocaine in blood plasma was collected and used for ML training and validation. After training, the reliability of proposed approach was tested on independently prepared samples, with unknown for SERS-ML cocaine presence or absence. As a result, the possibility of rapid determination of cocaine in blood plasma with a probability above 99.5% for cocaine concentrations up to 10-14 M was confirmed. Therefore, it is evident that the proposed approach has the ability to detect trace amounts of cocaine in bioliquids in an express and simple manner.

Pretreatment-free SERS sensing of microplastics using a self-attention-based neural network on hierarchically porous Ag foams

Low-cost detection systems are needed for the identification of microplastics (MPs) in environmental samples. However, their rapid identification is hindered by the need for complex isolation and pre-treatment methods. This study describes a comprehensive sensing platform to identify MPs in environmental samples without requiring independent separation or pre-treatment protocols. It leverages the physicochemical properties of macroporous-mesoporous silver (Ag) substrates templated with self-assembled polymeric micelles to concurrently separate and analyze multiple MP targets using surface-enhanced Raman spectroscopy (SERS). The hydrophobic layer on Ag aids in stabilizing the nanostructures in the environment and mitigates biofouling. To monitor complex samples with multiple MPs and to demultiplex numerous overlapping patterns, we develop a neural network (NN) algorithm called SpecATNet that employs a self-attention mechanism to resolve the complex dependencies and patterns in SERS data to identify six common types of MPs: polystyrene, polyethylene, polymethylmethacrylate, polytetrafluoroethylene, nylon, and polyethylene terephthalate. SpecATNet uses multi-label classification to analyze multi-component mixtures even in the presence of various interference agents. The combination of macroporous-mesoporous Ag substrates and self-attention-based NN technology holds potential to enable field monitoring of MPs by generating rich datasets that machines can interpret and analyze.

Surface-Enhanced Raman Spectroscopy and Artificial Neural Networks for Detection of MXene Flakes' Surface Terminations

The properties of MXene flakes, a new class of two-dimensional materials, are strictly determined by their surface termination. The most common termination groups are oxygen-containing (=O or -OH) and fluorine (-F), and their relative ratio is closely related to flake stability and catalytic activity. The surface termination can vary significantly among MXene flakes depending on the preparation route and is commonly determined after flake preparation by using X-ray photoelectron spectroscopy (XPS). In this paper, as an alternative approach, we propose the combination of surface-enhanced Raman spectroscopy (SERS) and artificial neural networks (ANN) for the precise and reliable determination of MXene flakes' (Ti3C2Tx) surface chemistry. Ti3C2Tx flakes were independently prepared by three scientific groups and subsequently measured using three different Raman spectrometers, employing resonant excitation wavelengths. Manual analysis of the SERS spectra did not enable accurate determination of the flake surface termination. However, the combined SERS-ANN approach allowed us to determine the surface termination with a high accuracy. The reliability of the method was verified by using a series of independently prepared samples. We also paid special attention to how the results of the SERS-ANN method are affected by the flake stability and differences in the conditions of flake preparation and Raman measurements. This way, we have developed a universal technique that is independent of the above-mentioned parameters, providing the results with accuracy similar to XPS, but enhanced in terms of analysis time and simplicity.

PCR-Free, Label-Free, and Centrifugation-Free Diagnosis of Multiplex Antibiotic Resistance Genes by Combining mDNA-Au@Fe3O4 from Heating Dry and DNA Concatamers with G-Triplex

Accurate identification of antibiotic resistance genes (ARGs) is crucial for improving treatment and controlling the spread of antibiotic-resistant bacteria (ARB). Herein, a novel PCR-free, centrifugation-free, and label-free magnetic fluorescent biosensor (MFB) was developed by combining polyA-medium DNA-polyT (mDNA, which contained a partial sequence of a target DNA), gold nanoparticle (AuNP)-anchored magnetic nanoparticle (Au@Fe3O4), complementary strand DNA (CS) of the target DNA, DNA concatamer with G-triplex (G3), and thioflavin T (ThT). Thereinto, Au@Fe3O4 nanoparticles were first capped by mDNA strands within 20 min using a simple hot drying method, and then CS was added and hybridized with mDNA on Au@Fe3O4. Second, a DNA concatamer was used to bind with CS on Au@Fe3O4. When an ARG was present in the sample, the CS would recognize it and release the DNA concatamer into solution by a toehold-mediated strand displacement reaction. Finally, under magnetic separation, the free DNA concatamers with G3 were taken out easily and bound with ThT, resulting in strong fluorescence signals. The fluorescence intensity of ThT was positively correlated with the concentration of the ARG. The whole analysis was accomplished within 1.5 h using 96-well plates. Remarkably, our MFB was universal; eight ARGs were detected by replacing the corresponding mDNA and CS in this study. To verify the practicability of our method, 12 clinically isolated strains were analyzed. The results of the MFB method were in good agreement with those of the quantitative real-time PCR method with an area under the curve of 0.92 (95% confidence interval: 0.8479 to 0.9932), sensitivity of 92.00%, and specificity of 91.55%. Above all, the MFB assay established here is simple, low-cost, and universal and has great potential for applications in the identification of ARGs.

Bioinspired superhydrophobic SERS substrates for machine learning assisted miRNA detection in complex biomatrix below femtomolar limit

Surface-enhanced Raman spectroscopy (SERS) is an analytical method with high potential in the field of medicine. The design of SERS substrates, based on specific morphology and/or chemical modification, allow the recognition of the presence of specific analytes with precision close to a single-molecule detection limit. However, the SERS analysis of real samples is significantly complicated by the presence of a large number of "minor" molecules that can shield the signal from the target analyte and make it impossible to determine it in practice. In this work, an advanced SERS approach was used for the detection of cancer-related miRNA-21 in blood plasma, used as a molecular model background. The approach was based on the combination of the biomimetic plasmon-active SERS substrate, its tuned surface chemistry and advanced SERS data analysis, making use of artificial machine learning. In the first step, biomimetic SERS substrates were created using a butterfly wing as a starting template. The substrates were covered by thin Au layer and covalently grafted with hydrophobic chemical moieties to introduce superhydrophobic and water-adhesive properties. The self-concentration of the analyte on the substrates was achieved by minimizing the contact area between the analyte drop and the substrate, which is facilitated by surface superhydrophobicity and additionally enhanced by drop evaporation on the flipped over substrate. Due to the presence of cancer miRNA and blood plasma background, the measured SERS spectra represent a complex of interfering peaks. Thus, their interpretation was carried out using a specially trained machine learning model. As a result, reliable and repeatable quantitative detection of miRNAs below the femtomolar level (up to 10-16 M) on the background of human blood plasma becomes possible.

Unraveling surface-enhanced Raman spectroscopy results through chemometrics and machine learning: principles, progress, and trends

Surface-enhanced Raman spectroscopy (SERS) has gained increasing attention because it provides rich chemical information and high sensitivity, being applicable in many scientific fields including medical diagnosis, forensic analysis, food control, and microbiology. Although SERS is often limited by the lack of selectivity in the analysis of samples with complex matrices, the use of multivariate statistics and mathematical tools has been demonstrated to be an efficient strategy to circumvent this issue. Importantly, since the rapid development of artificial intelligence has been promoting the implementation of a wide variety of advanced multivariate methods in SERS, a discussion about the extent of their synergy and possible standardization becomes necessary. This critical review comprises the principles, advantages, and limitations of coupling SERS with chemometrics and machine learning for both qualitative and quantitative analytical applications. Recent advances and trends in combining SERS with uncommonly used but powerful data analysis tools are also discussed. Finally, a section on benchmarking and tips for selecting the suitable chemometric/machine learning method is included. We believe this will help to move SERS from an alternative detection strategy to a general analytical technique for real-life applications.

Classification method based on Siamese-like neural network for inter-species blood Raman spectra similarity measure

Analysis of blood species is an extremely important part in customs inspection, forensic investigation, wildlife protection and other fields. In this study, a classification method based on Siamese-like neural network (SNN) for interspecies blood (22 species) was proposed to measure Raman Spectra similarity. The average accuracy was above 99.20% in the test set of spectra (known species) that did not appear in the training set. This model could detect species not represented in the dataset underlying the model. After adding new species to the training set, we can update the training based on the original model without retraining the model from scratch. For species with lower accuracy, SNN model can be trained intensively in the form of enriched training data for that species. A single model can achieve both multiple-classification and binary classification functions. Moreover, SNN showed higher accuracy rates when trained with smaller datasets compared to other methods.

Deep learning approach to overcome signal fluctuations in SERS for efficient On-Site trace explosives detection

Surface-enhanced Raman spectroscopy (SERS) is an improved Raman spectroscopy technique to identify the analyte under study uniquely. At the laboratory scale, SERS has realised a huge potential to detect trace analytes with promising applications across multiple disciplines. However, onsite detection with SERS is still limited, given the unwanted glitches of signal reliability and blinking. SERS has inherent signal fluctuations due to multiple factors such as analyte adsorption, inhomogeneous distribution of hotspots, molecule orientation etc. making it a stochastic process. Given these signal fluctuations, validating a signal as a representation of the analyte often relies on an expert's knowledge. Here we present a neural network-aided SERS model (NNAS) without expert interference to efficiently identify reliable SERS spectra of trace explosives (tetryl and picric acid) and a dye molecule (crystal violet). The model uses the signal-to-noise ratio approach to label the spectra as representative (RS) and non-representative (NRS), eliminating the reliability of the expert. Further, experimental conditions were systematically varied to simulate general variations in SERS instrumentation, and a deep-learning model was trained. The model has been validated with a validation set followed by out-of-sample testing with an accuracy of 98% for all the analytes. We believe this model can efficiently bridge the gap between laboratory and on-site detection using SERS.

Cheminformatics Study on Structural and Bactericidal Activity of Latest Generation β-Lactams on Widespread Pathogens

Raman spectra of oxacillin (OXN), carbenicillin (CBC), and azlocillin (AZL) are reported for the first time together with their full assignment of the normal modes, as calculated using Density Functional Theory (DFT) methods with the B3LYP exchange-correlation functional coupled to the 6-31G(d) and 6-311+G(2d,p) basis sets. Molecular docking studies were performed on five penicillins, including OXN, CBC, and AZL. Subsequently, their chemical reactivity and correlated efficiency towards specific pathogenic strains were revealed by combining frontier molecular orbital (FMO) data with molecular electrostatic potential (MEP) surfaces. Their bactericidal activity was tested and confirmed on a couple of species, both Gram-positive and Gram-negative, by using the disk diffusion method. Additionally, a surface-enhanced Raman spectroscopy (SERS)-principal component analysis (PCA)-based resistogram of A. hydrophila is proposed as a clinically relevant insight resulting from the synergistic cheminformatics and vibrational study on CBC and AZL.

Temperature-Responsive Nanocarrier-Regulated Alternative Release of "Cargos" for a Multiplex Photoelectrochemical Bioassay of Antibiotic-Resistant Genes

A smart temperature stimuli-driven multiplex photoelectrochemical (PEC) assay was constructed for antibiotic resistance genes (ARGs) detection, where the stimuli-responsive gatekeeping by regulating the alternative release of "cargo" allowed for the simultaneous detection of multiple tetracycline resistance gene, using tetA (T_DNA1) and tetC (T_DNA2) as the model. Dual temperature-responsive nanoassemblies were embedded in the PEC bioassay as signal DNA tages: one thermoresponsive polymer (poly(N-isopropylacrylamide), PNIPAM)-capped mesoporous silica nanoparticles (MSN) with loading the "cargo" of HgO nanoparticles as signal DNA1 tags (S_DNA1-PNIPAM@MSN@HgONPs) and the other antimony tartrate (SbT)-anchored silica nanospheres as signal DNA2 tags (S_DNA2-SbT@SiO₂NSs). At 20 °C, below the lower critical solution temperature (LCST) of PNIPAM, the "gatekeeper" PNIPAM in S_DNA1-PNIPAM@MSN@HgONPs was in an ON state, igniting Hg²⁺ release through the pore of SiO₂. While at above LCST (40 °C), it was in an OFF state. Likewise, the thermo-dependent dissociation of SbT endowed the grafted S_DNA2 tags switching from the OFF (at 20 °C) to ON state (at 40 °C), igniting SbO⁺ release. The released Hg²⁺ and SbO⁺ triggered the amplified photocurrents due to the structure evolution of the photoactive layer into HgS/ZnS or Sb₂S₃/ZnS heterostructure, thus achieving sensitive detection of multiple ARGs: tetA, tetC, tetG, tetM, tetO, tetZ, tetX, and tetW. Combined with heat map analysis, rapid screening of the ARGs profiles in 12 samples could be realized. This bioassay is simple and accessible for multiple genes analysis with the detection limit down to 0.50 nM. And it was successfully applied for measuring tetracycline ARGs in real sludge samples.

New Trends in Nanoarchitectured SERS Substrates: Nanospaces, 2D Materials, and Organic Heterostructures

This article reviews recent fabrication methods for surface-enhanced Raman spectroscopy (SERS) substrates with a focus on advanced nanoarchitecture based on noble metals with special nanospaces (round tips, gaps, and porous spaces), nanolayered 2D materials, including hybridization with metallic nanostructures (NSs), and the contemporary repertoire of nanoarchitecturing with organic molecules. The use of SERS for multidisciplinary applications has been extensively investigated because the considerably enhanced signal intensity enables the detection of a very small number of molecules with molecular fingerprints. Nanoarchitecture strategies for the design of new NSs play a vital role in developing SERS substrates. In this review, recent achievements with respect to the special morphology of metallic NSs are discussed, and future directions are outlined for the development of available NSs with reproducible preparation and well-controlled nanoarchitecture. Nanolayered 2D materials are proposed for SERS applications as an alternative to the noble metals. The modern solutions to existing limitations for their applications are described together with the state-of-the-art in bio/environmental SERS sensing using 2D materials-based composites. To complement the existing toolbox of plasmonic inorganic NSs, hybridization with organic molecules is proposed to improve the stability of NSs and selectivity of SERS sensing by hybridizing with small or large organic molecules.

Raman spectrum matching with contrastive representation learning

An effective contrastive representation learning method for spectra identification with a frequentist guarantee of including the correct class prediction on two Raman datasets (Mineral and Organic) and one SERS dataset (Bacteria).