Quantitative H NMR (H qNMR) is an ideal tool for bioprocess monitoring because it can comprehensively detect and quantify diverse metabolites that significantly influence bioprocess performance. However, the long experiment time associated with the H qNMR, due to the long longitudinal relaxation time (T1) of some metabolites, does not meet the requirements for high-throughput analysis. We developed a high-throughput H qNMR method for bioprocess analysis using a short relaxation delay (D1) to reduce analytical time and a correction factor (k) to compensate for incomplete relaxation. A total of 27 metabolites were quantified using spectral deconvolution via a peak fitting algorithm and MCR-ALS. Methodological validation results indicated that the precision and accuracy of the developed qNMR method were consistently high across different D1 values, with LOQs ranging from 0.008 to 0.13 mM and LODs ranging from 0.024 to 0.38 mM. Notably, a longer D1 value generally resulted in lower LODs and LOQs for most metabolites. A D1 value of 4 s was optimal for balancing analysis time and performance. The method is broadly applicable for bioprocess monitoring and control, offering valuable guidance for optimizing CHO cell culture processes and improving yield.

G-quadruplex/hemin DNAzyme, a versatile tool for biosensing, is challenged by its low peroxidase-mimic activities. The addition of NH may offer an efficient approach to improve its activity. However, the detailed impact of NH on its catalytic activity remains unclear, confusing the selection of appropriate DNAzymes for biosensing applications. Here, we conducted a comprehensive examination of the influence of NH on G-quadruplex/hemin DNAzyme. The results revealed that all DNAzymes with different G-quadruplex topologies exhibited increased catalytic activities in the presence of NH relative to K, followed by the subsequent activity order: parallel > hybrid > antiparallel. Further investigations indicated that the increased catalytic activity can be ascribed to the increased stability of the G-quadruplex/hemin complex, elevated reaction velocity, and improved substrate affinity. Leveraging the significant disparity in enzymatic activity between parallel and antiparallel G-quadruplexes, an allosteric sensor based on the Pb-induced topological conformation was developed for sensitive detection of Pb in the NH-boosted G-quadruplex/hemin DNAzyme system (LOD, 1.56 nM), indicating potential for practical applications. Our discovery improves the understanding of NH-boosted G-quadruplex/hemin DNAzyme and may facilitate the development of biosensors.

Data-independent acquisition (DIA) and data-dependent acquisition (DDA) are frequently employed in the execution of tandem mass spectrometry (MS2) analyses. This study explored the application of DIA (MSe) and DDA (fast-DDA) in liquid chromatography-mass spectrometry (LC-MS)-based untargeted metabolomics using Panax genus samples. MSe provided comprehensive sample information, extracting more ion peaks with better peak shape and increased scan points compared to fast-DDA. Features from MSe data are four times more than those from fast-DDA data. Fast-DDA, however, delivered high-quality MS2 data, enhancing compound annotation via the GNPS web tool. Database matches with fast-DDA data were nearly 35 times greater than those using MSe data. Therefore, combining MSe and fast-DDA can improve data analysis and metabolite annotation. An improved workflow integrating DIA and DDA was proposed, requiring additional QC sample injections for DDA analysis but eliminating the need for sample reprocessing and re-analysis, thus saving time and resources. The established workflow was applied to the Panax genus samples analysis to confirm its applicability. This study offers a deeper understanding of DIA and DDA, guiding the selection of data acquisition strategies for LC-MS-based untargeted metabolomics.

A hemicyanine-based colorimetric and fluorometric sensor, 2-(2-(2,3,5,6,8,9-hexahydrobenzo[b][1,4,7,10]tetraoxacyclododecin-12-yl)vinyl)-3,3-dimethyl-1-propyl-3H-indol-1-ium iodide (MH-5), was developed and synthesized to detect Li and CN ions in DMSO-PBS buffer solution (10 mM, pH 7.25, v/v 1:9). MH-5 displayed a rapid and highly selective colorimetric response to both Li and CN, indicated by a distinct color change from pink to pale pink in the presence of Li and to colorless upon CN detection, without interference from other cations or anions. The interaction mechanisms of MH-5 with Li and CN ions were investigated using various analytical techniques, including H NMR, ESI-MS, FT-IR spectroscopy, and Job's plot analysis. These studies suggest that CN is detected through nucleophilic addition to the indolium moiety of MH-5, while Li detection occurs via coordination with oxygen atoms in the crown ether structure. The fluorescence-based detection limits for Li and CN were determined to be 0.150 µM and 0.154 µM, respectively. Additionally, MH-5 was evaluated in living cells, demonstrating effective cell penetration and reliable detection of Li and CN ions for potential bio-imaging applications.

Fluorescence biosensors hold significant importance for testing antibiotic residues which seriously endanger public health. However, how to adopt appropriate strategies to address the false result disadvantage involved in traditional single-channel biosensors is still a great challenge. Meanwhile, too much attention focused on designing signal amplification strategies of biosensors unavoidably decreases their detection efficiency. Herein, we combined the designed dual DNA recycling amplification strategy with CRISPR/Cas12a-mediated dual-channel signal output mode to successfully develop a novel ratiometric fluorescence biosensor for testing kanamycin (Kana) residues in complex sample matrices. The first recycling was formed from an exonuclease-assisted aptamer recognition reaction, which also triggered another cascade DNA recycling to amplify the release of the Cas12a activator. With the non-discrimination cleavage of Cas12a to cause reverse fluorescence changes of copper nanoclusters and an AMAC-labeled signal DNA, the ratiometric signal transduction strategy was constructed. Under optimal conditions, this biosensor could be applied for ultrasensitive testing of Kana antibiotics in a five-order of magnitude wide linear range with a low detection limit of 17.2 fg mL. Benefiting from the self-correction function of the ratiometric signal transduction mode, it showed promising practicality in lake water and milk samples with the relative error less than 4.9% to the standard ELISA results. Besides CRISPR/Cas12a-based fluorescence output efficiency improvement, this biosensor also excluded the complicated manipulations and expensive instruments required in traditional methods. Therefore, it provides a good choice for expanding the application of fluorescence biosensing technology for practical analysis application.

Polyesters comprise the greatest proportion of textile fibers and are found in various everyday goods; hence, polyester fibers are a significant source of microplastic pollution and textile waste. The specific chemical composition of commercial polyester fibers is often proprietary and mostly assumed to be poly(ethylene terephthalate) (PET). Polyester is a class of polymers that include poly(butylene terephthalate) (PBT), poly(cyclohexylenedimethylene terephthalate) (PCT), and poly(ethylene naphthalate) (PEN), as well as biodegradable polymers. Our study aims to clarify whether household polyester products are primarily PET, are labeled accurately, or contain phthalate additives by applying double-shot pyrolysis-gas chromatography/mass spectroscopy (Py-GC/MS). We analyzed four scientific-grade polyester reference standards, 52 manufacturer-grade polyester fibers or pellets, and 229 samples from 193 consumer polyester products. From the pyrograms, samples were predominantly identified as PET (87.4%, 95% CI [93.5-81.3%]), but five samples were identified as a different polyester, nine as non-polyester polymers, and 23 as a blend of PET with another polymer. From the thermal desorption chromatograms, diethyl phthalate was the most frequently detected phthalate, found in 23.3% (95% CI [17.3-29.3%]) of the consumer products, including children's toys. Double-shot py-GC/MS advantageously results in these empirical data that (1) counter the assumption that products labeled polyester are always PET, (2) emphasize the importance of creating spectral libraries with well-characterized materials for accurate polymer identification of unknown plastic particles, and (3) demonstrate that phthalates are common additives in household products.

Comprehensive characterization of petroleum-derived sulfur compounds is crucial for researching and developing desulfurization catalysts, equipment, and processes. However, the complex composition and low concentration of sulfur compounds in oils make it challenging for molecular-level separation and characterization. In this work, sulfur compounds in the straight-run diesel fraction were selectively separated from the oil by the methylation/demethylation method, effectively yielding high-purity thiophenes and sulfides. Molecular-level compositional and structural characterization of sulfur compounds was accomplished through comprehensive two-dimensional gas chromatography (GC × GC) coupled with time-of-flight mass spectrometry. The separation significantly enhances the detection sensitivity for low-content sulfur compounds, thereby enabling a more comprehensive characterization of their molecular compositions and structures. Sulfur compounds with diverse skeletons and carbon numbers were tentatively characterized, including 1~3 cyclic sulfides, thiophenes, benzothiophenes, dihydrobenzothiophenes, dibenzothiophenes, tetrahydrodibenzothiophenes, phenanthrothiophenes, and benzonaphthothiophenes. Additionally, hundreds of individual sulfur compounds were characterized by mass spectrometry. Quantitative analysis for individual compounds and compound types was conducted using a flame ionization detector.

Vitamin C (L-( +)-ascorbic acid, L-AA) is an essential micronutrient. Its diastereoisomer, D-(-)-isoascorbic acid (D-IAA), is only 5% as active as L-AA. Therefore, it is crucial to identify and characterize the diastereoisomers of ascorbic acid. In this work, a straightforward and direct method using ion mobility-mass spectrometry (IM-MS) was proposed to identify ascorbic acid isomers. Ternary complexes [L-AA + γ-CD + 2Cs-H] and [D-IAA + γ-CD + 2Cs-H], formed by non-covalent interaction of the isomer with the selective agent γ-CD and the metal ion Cs, were separated in ion drift tubes with a resolution of R as high as 1.398. Meanwhile, comparisons with different CDs and metal ions revealed varying separation efficiencies. Theoretical calculations were conducted to determine the optimal conformations of [L-AA + γ-CD + 2Cs-H] and [D-IAA + γ-CD + 2Cs-H]. Conformational analysis highlighted distinct structural differences at the molecular level, providing insight into the mobility separation mechanism of AA via the formation of ternary complexes with γ-CD and metal ions. Additionally, a quantitative analysis for the determination of chiral isomers was established with effective linearity and acceptable sensitivity. The method was successfully applied to assess L-AA/D-IAA content in pharmaceuticals and fruit samples.

The most powerful and widely used detector for clinical and toxicological analyses is undoubtedly the mass spectrometer (MS) due to its specificity and high sensitivity. However, it cannot differentiate enantiomers from each other, as well as cannot easily distinguish between positional isomers. Therefore, the components of interest to the analysis at hand need to be separated prior to their detection by MS in order to reliably identify and quantify their enantiomers and positional isomers. In the present study, the simultaneous chemo- and enantioseparation of the drugs of abuse, 2-, 3-, and 4-methylmethcathinones, is described. The developed method was applied to 15 oral fluid (OF) samples collected by police in Belgium and found positive for mephedrone (4-methylmethcathinone, 4-MMC) based on a nonselective method for enantiomers and positional isomers. However, re-analyzing these samples with the analytical method proposed in this report indicated that mephedrone was present in only 1 of them, while 6 samples contained 3-methylmethcathinone (3-MMC) and 12 samples contained 2-methylmethcathinone (2-MMC). At the same time, four OF samples contained both 2- and 3-MMC. The developed method enabled the enantioselective analysis of major metabolites of methylmethcathinones, such as their N-demethyl derivatives (nor-MMCs), as well as, at least partially, of their dihydrometabolites. In addition to their positional isomers, other structural isomers of MMCs, such as buphedrone, metamfepramone, and ethcathinone, could also be detected enantioselectively.

As a synthetic polymer, methoxy-polyethylene glycol propionic acid (M-PEG-PA) is widely used in the biomedicine field. Unraveling the pharmacokinetic behavior of M-PEG-PA in vivo is crucial for evaluating the safety and efficiency of M-PEG-PA-related polymers or drug delivery systems. A high-throughput and green ultra-performance convergence chromatography-tandem mass spectrometry (UPC-MS/MS) assay was firstly developed and validated for the determination of M-PEG-PA polymers in a biological matrix. The MRM transition (mass to charge ratio, precursor ions→fragment ions) of m/z 367.2→118.9 was used to quantify M-PEG-PA in this study. The throughput of the assay is high and the total running time for each sample was only 2 min. The linear range of the developed UPC-MS/MS assay for quantification of M-PEG-PA in a biological matrix is 0.05 to 30 μg/mL (R≥0.995). Intra-day and inter-day precisions for the determination of M-PEG-PA by this analytical assay were <6.99%. The absolute recoveries and matrix effects of M-PEG-PA ranged from 79.50 to 92.47% and 68.72 to 81.73%, respectively. The UPC-MS/MS assay was successfully applied to quantify the concentrations of M-PEG-PA polymers in rat plasma samples.

This work is based on the development and optimization of a pressurized liquid extraction method to obtain extracts from peanut shells with the highest possible amount/number of bioactive compounds, mainly flavonoids, with senolytic activity and antioxidant capacity. To achieve optimal extraction conditions, a design of experiments approach was employed to perform a limited and relatively reduced number of experiments. The extracts were consecutively analyzed by methods adapted to the peanut shell matrix to determine antioxidant capacity, total flavonoids, and total phenolic compounds. Additionally, a high-performance liquid chromatography coupled with diode array detection method was developed and validated to quantify individual phenolic compounds, with confirmation provided by mass spectrometry. Moreover, amino acid profiling was performed using gas chromatography coupled with mass spectrometry. Finally, the optimized extraction conditions and analytical methods were applied to analyze six commercial peanut shell samples. The results indicate that the optimized pressurized liquid extraction method using ethanol effectively extracts substantial amounts of bioactive compounds, especially flavonoids, which have broad applications across different industries. This contributes to a strategic valorization approach that promotes a Circular Economy.

Cyanobacteria are prokaryotic organisms that can form large monospecific blooms, which pose a risk to human and animal health as some species produce toxic secondary metabolites called cyanotoxins. Multiclass cyanotoxin analysis is challenging due to varying chemical and physical properties between classes, as well as potentially large numbers of analogues within each class. Incorporating anatoxins (ATXs) into multiclass methods can be particularly challenging due to their small molecular size, potential interferences, polarity, and a lack of chemical standards for most analogues. Here, we present the development of a multiclass LC-MS/MS method and a quantitative calibration solution for aetokthonotoxin (AETX), an emerging cyanotoxin linked to mass mortalities of bald eagles in the Eastern United States. The developed method is capable of detecting 17 microcystins (MCs), nodularin-R, three cylindrospermopsins (CYNs), AETX, and 17 ATXs, including recently tentatively identified 10-hydroxy analogues. Analytes were identified by retention time and product ion ratio matching with available standards. The method was evaluated with respect to limits of detection (LODs), linear range, accuracy, and precision using neat and matrix matched standards. LODs in wet cyanobacterial biofilms ranged from 0.14 ng/g for CYN to 2.8 ng/g for [Dha]MC-LR with accuracies ranging from 65% for [Leu]MC-LY to 116% for CYN. Finally, the method's application was demonstrated through analysis of cyanobacterial field samples, a dietary supplement matrix reference material, and passive sampler extracts to assess versatility within different matrices.

The complexity of blood and its matrix in high-performance liquid chromatography (HPLC) glycated hemoglobin (HbA1c) assays is closely related to column performance. However, the available prefilters with a single structure or surface membrane materials are not ideal for column protection, and coupled with the complexity of blood samples, leads to rapid degradation of column performance. Therefore, we have developed a new microprefilter with a three-stage filtration design and depth filter material to protect the column. All filter materials used in the preparation of microprefilters were characterized, screened, and optimized, and then manufactured on the basis of optimized filter materials, which are depth filter material microprefilters. Based on the material and structural design, microprefilters were capable of filtering particulate matter from test samples on a step-by-step basis to avoid the plugging effect that occurs when all sizes of substances are gathered together. Moreover, all newly developed microprefilters can be tested more times, up to 600 times. Microprefilters with small-pore-size final filtration membranes of polyethersulfone, hydrophilic polytetrafluoroethylene, and mixed cellulose showed excellent column protection in terms of column efficiency, HbA1c retention time, number of column tests, and column backpressure, and prolonged column lifetime by as much as 20-30% compared with microprefilters with large-pore-size final membranes. Our study provides valuable depth filter material microprefilters with multistage filtration for chromatography columns, and showed excellent column protection and prolonged column lifetime. Meanwhile, microprefilters can be tested more times. The newly developed microprefilters with a small-pore-size final membrane are the optimal choice for column protection of the HbA1c assay.

Multicellular tumor spheroids (MCTSs) play an important role in biological studies and cancer research. There is an emerging research interest in molecular profiling and drug distribution of MCTSs by leveraging the superior sensitivity and molecular specificity of mass spectrometry imaging (MSI). Current methods for sample preparation of MCTSs can suffer from low throughput, as MCTSs are typically individually transferred from cell culture into an MSI embedding media and sectioned individually, or sometimes, a few spheroids are placed in a small block of embedding media in preparation for MSI. Here, we developed a method to minimize the sample preparation steps needed to create high-throughput MCTS frozen sections for MSI. Agarose-based microarrays created from Microtissues molds were used during MCTS culturing, after which the entire MCTS agarose microarray was taken out of the cell culture well and then directly embedded in 5% gelatin, without the need for a transfer step for each individual MCTS into the embedding media. This method enables rapid profiling of up to 81 MCTSs for larger MCTSs (500-800 µm) or up to 256 MCTSs for smaller MCTSs (200-300 µm) in a single section, remarkably improving the throughput possible for MSI MCTS workflows. Notably, sectioning MCTSs together in the agarose microarray also improves MCTS visualization during sectioning, such that staining each MCTS section to ensure the presence of the MCTSs within the embedding media is not necessary during the sectioning process. The method described here provides a more direct, convenient strategy to achieve high-throughput sections. MSI MCTS sectioning throughput is an important advancement for both pharmaceutical testing of MCTS; the direct transfer 3D cell cultures grown within cell culture-compatible polymer scaffolding are also critical for expanding MSI for the characterization of microfluidic and complex in vitro models, where agarose is readily utilized as a non-adhesive 3D cell culture scaffold.

Diabetes mellitus is a metabolic disorder that impacts millions of individuals globally. In the treatment of this condition, it is imperative to explore natural resources for therapeutic agents that exhibit fewer adverse effects and enhanced efficacy. Currently, the methods employed for isolating anti-diabetic lead compounds from natural sources are often intricate and time-consuming. Therefore, there is an urgent need to develop efficient and rapid screening techniques. In this study, α-amylase was immobilized using a novel polydopamine/L-cysteine bifunctionalized magnetic mesoporous silica composite material (FeO@nSiO@mSiO@PDA@L-Cys) for the first time. A ligand fishing approach utilizing the immobilized α-amylase was developed to rapidly screen for α-amylase inhibitors from Aloe vera. Characterization and property analysis of the immobilized enzyme showed that the immobilized α-amylase exhibited exceptional stability and reusability. Two ligands were successfully screened from Aloe vera and then characterized as aloin B and aloin A using ultra-high performance liquid chromatography tandem mass spectrometry. Their respective IC values were 0.99 ± 0.09 mM and 1.14 ± 0.05 mM. Molecular docking studies confirmed the interaction of both ligands with specific amino acid residues within the active site of α-amylase. The study presents a fast and efficient approach for screening α-amylase inhibitors from intricate natural sources, thereby offering significant potential for the development of anti-diabetic agents.

Anaerobic gut fungi (AGF) have emerged as promising candidates for optimized biogas and biofuel production due to their unique repertoire of potent lignocellulose-degrading enzymes. However, identifying AGF strains through standard fungal DNA barcodes still poses challenges due to their distinct genomic features. This study explored the applicability of matrix-assisted laser desorption ionization time-of-flight mass spectrometry (MALDI) and direct analysis in real-time (DART) mass spectrometry (MS) as alternative methods for AGF identification. Further, the capability of the methods to differentiate strains from different growth phases was investigated. The study found that both MALDI and DART were viable methods for AGF strain identification. MALDI proved to be a precise and robust technique for strain discrimination with prediction accuracies of 94% for unknown standard samples. Even at longer growth times (>3 weeks) MALDI achieved good prediction accuracies with 84%; however, younger cultures (72 h) were only predicted with 63% accuracy. The fast on-target lysis with minimal chemical demand yielded suitable spectra for strain differentiation. DART MS, while effective with prediction accuracies of samples with the same age of up to 93%, exhibited lower prediction accuracies for cultures of different ages, with 14% for young (72 h) and 71% for old (>3 weeks) samples. Further research could enhance the capabilities of these mass spectrometry methods for AGF identification and broaden their application to species-level discrimination and a wider range of AGF genera.

Dual-signal mode sensors that can self-validate detection results have attracted considerable interest; however, creating those with superior overall performance still presents significant challenges. Herein, we develop a unique photoelectrochemical (PEC) and photothermal (PT) dual-mode biosensor targeting microRNA-221 (miRNA-221), built on an innovative entropy-driven DNA circuit (EDC). The zinc oxide nanorods (ZnO NRs) serve as PEC beacons, while copper sulfide nanoparticles (CuS NPs) function as photocurrent inhibitors and PT beacons, both biofunctionalized with DNAs before being assembled through partial base pairing. When target miRNA-221 is present, the EDC activates and releases output DNAs that open partially hybridized strands anchored to ZnO NRs via competitive assembly. This process liberates CuS-DNA1 and restores the suppressed photocurrent. The results demonstrate linear relationships between photocurrent/temperature increment and the logarithm of target concentration across ranges of 1.0 fmol L-50.0 pmol L (limit of detection (LOD): 0.35 fmol L) and 5.0×10 fmol L-5.0 nmol L (LOD: 1.22×10 fmol L), respectively. Compared to conventional EDCs, our optimally designed EDC not only doubles the output DNA yield but also significantly enhances sensor sensitivity. Additionally, the target-triggered EDC amplification strategy effectively minimizes reversibility in each reaction step, preserves base sequence integrity, boosts efficiency, and demonstrates strong thermal stability and selectivity, thereby increasing the specificity of the dual-mode biosensor. Furthermore, ZnO NR photoelectric beacons fabricated via electrodeposition greatly improve the stability and controllability of the photoelectrode while avoiding lengthy modification processes. Overall, this thoughtfully engineered dual-mode biosensor offers numerous advantages, including a wide linear range, excellent stability, high reproducibility, and user-friendly operation. Specifically, this signal-on type dual-signal output biosensor enables self-confirmation of detection results, significantly enhancing both accuracy and reliability.

Organophosphonates (OPs) are widely used as chelating agents in domestic and industrial applications. While regarded as hardly biodegradable, OPs can undergo abiotic transformation with phosphate (PO) as a main transformation product. As some OPs are suspected precursors of glyphosate in surface waters, their environmental fate is of current interest. Due to analytical challenges posed by quantification of individual OPs, monitoring PO formation is a widely used proxy to monitor OP transformations. The molybdenum blue (MB) method, employing UV/Vis spectroscopy, is frequently used for PO quantification due to its sensitivity and operational simplicity. However, while interference of certain inorganic ions is well-documented, the effects of OPs on the accuracy of the MB method remain unexplored. This study investigated the effects of six OPs, namely N-(phosphonomethyl)glycine (glyphosate), 1-hydroxyethylidene(1,1-diphosphonic acid) (HEDP), iminodi(methylene phosphonate) (IDMP), aminotris(methylene phosphonate) (ATMP), ethylenediaminetetra(methylene phosphonate) (EDTMP), and diethylenetriaminepenta(methylene phosphonate) (DTPMP). Spectral analysis of pure PO standards using the MB method exhibits two characteristic absorption maxima (λ) at 710 and 880 nm. In the presence of OPs, a new λ appears around 760 nm. This is accompanied by an increase in absorbance values at both 710 and 880 nm, leading to significant over-quantification of PO concentrations. Among the evaluated OPs, DTPMP exhibits the most substantial interference (PO over-quantification by up to 240%), while glyphosate causes minimal interference (≤ 20%). The effects are most pronounced at OPs:PO ratios ≥1. A case study simulating DTPMP transformation confirms PO over-quantification of up to 350%, revealing limitations of the MB method. Therefore, careful data evaluation and complementary analytical techniques for accurate PO measurements are indispensable in OP transformation research.

Contamination of surface water and drinking water with pharmaceuticals presents an environmental concern. It has been shown to affect aquatic organisms and have adverse health effects on humans. One of the most common pharmaceutical contaminants is the opioid analgesic tramadol. In this communication, we report on the first surface plasmon resonance biosensor-based detection of tramadol in water. The biosensor utilizes a binding inhibition format and enables detection of tramadol at a wide range of concentrations (5 orders of magnitude) with a limit of detection of 0.52 µg/L. The results of a small-scale environmental study are reported in which the biosensor was used to analyze river water samples. The results were found to agree well with those obtained using the liquid chromatography-tandem mass spectrometry (HPLC-MS/MS).

This study proposes a rapid identification method for foodborne pathogens by combining Raman spectroscopy with explainable machine learning. Spectral data of nine common foodborne pathogens are collected using a laser confocal Raman spectrometer, and their characteristic Raman peaks are identified and analyzed. Key spectral features are extracted using competitive adaptive reweighted sampling (CARS) and the successive projections algorithm (SPA), while t-distributed stochastic neighbor embedding (t-SNE) is employed for visualization. Subsequently, classification models, including support vector machine (SVM) and random forest (RF), are developed, and the optimal model is selected based on classification accuracy (ACC), with the RF model achieving a test accuracy of 98.91%. To enhance the interpretability of the model, Shapley Additive exPlanations (SHAP) analysis is applied to evaluate the contribution of each spectral feature to the classification results, identifying critical Raman shifts significantly influencing pathogen classification. The results demonstrate that CARS-SPA feature selection not only improves the accuracy and efficiency of the classification model but also enhances its transparency and reliability. This study optimizes the workflow for food safety testing, reduces the risk of foodborne diseases, and provides robust technical support for public health and safety.