Single-cell detection and linear discriminant analysis of bacterial Raman spectra in glass filter microholes

We report a study of micro-Raman spectroscopy towards an optimal approach for single cell measurements for the detection of bacteria by vibrational spectroscopy. The use of glass membrane filters was tested by microfiltration to separate individual bacterial cells. The glass membrane filters were applied to the study of Raman spectral classification analysis. This approach achieved the capture and individual detection of spiked bacterial cells. Linear discriminant analysis (LDA) of Raman spectra measured on glass membrane filters was successfully used to distinguish several bacterial species.

Gold Nanorod Amalgamation: Machine Learning Empowered Discrimination of Biothiol and Thiol Ratios

Biothiols, characterized by thiol groups, exhibit remarkable affinity for certain metals, playing pivotal roles in intracellular and extracellular biological processes. Fluctuations in their levels profoundly impact overall physiological health. Despite the development of various probes for biothiol detection and quantification, their inability to monitor thiol-to-disulfide state transitions persists as a limitation. Given their association with pathologies, early detection remains imperative. Gold nanorod (AuNR)-based colorimetric probes have garnered attention for their utility in visual diagnostic assays. Herein, we present a cost-effective, and sensitive multicolor ratio measuring probe enabling on-site simultaneous identification, discrimination, and quantification of essential biothiols─cysteine (CYS), glutathione (GSH), cystine (CYSS), and glutathione disulfide (GSSG)─while also quantifying thiol-to-disulfide ratios. Our investigation clarifies the probe's functionality, elucidating etching and antietching mechanisms based on sulfhydryl group coordination with Hg2+. This coordination impedes gold amalgam formation, facilitating discriminative detection via AuNR size and aspect ratio modulation, validated by transmission electron microscopy. Notably, distinct rainbow-like fingerprint patterns were discernible both visually and spectroscopically for the aforementioned biothiols and their respective thiol-to-disulfide ratios. Subsequent qualitative and quantitative analyses via linear discriminant analysis (LDA) and partial least squares regression revealed linear correlations over broad concentration ranges (CYS: 1.9-40 μmol L-1, GSH: 3.2-200.0 μmol L-1, CYSS: 2.0-70.0 μmol L-1, GSSG: 3.7-100.0 μmol L-1), with detection limits of 0.66 μmol L-1 (CYS), 1.07 μmol L-1 (GSH), 0.69 μmol L-1 (CYSS), and 1.24 μmol L-1 (GSSG). Moreover, thiol-to-disulfide ratios exhibited linear patterns within 0.2-5 μmol L-1, with detection limits of 0.13 and 0.09 μmol L-1, and exceptional analytical sensitivities of 32.648 and 49.782 for (CYS/CYSS) and (GSH/GSSG), respectively. Lastly, we evaluated the probe's performance in complex matrices relative to aqueous media, both quantitatively and qualitatively.

Automated collective variable discovery for MFSD2A transporter from molecular dynamics simulations

Biomolecules often exhibit complex free energy landscapes in which long-lived metastable states are separated by large energy barriers. Overcoming these barriers to robustly sample transitions between the metastable states with classical molecular dynamics (MD) simulations presents a challenge. To circumvent this issue, collective variable (CV)-based enhanced sampling MD approaches are often employed. Traditional CV selection relies on intuition and prior knowledge of the system. This approach introduces bias, which can lead to incomplete mechanistic insights. Thus, automated CV detection is desired to gain a deeper understanding of the system/process. Analysis of MD data with various machine-learning algorithms, such as principal component analysis (PCA), support vector machine, and linear discriminant analysis (LDA) based approaches have been implemented for automated CV detection. However, their performance has not been systematically evaluated on structurally and mechanistically complex biological systems. Here, we applied these methods to MD simulations of the MFSD2A (Major Facilitator Superfamily Domain 2A) lysolipid transporter in multiple functionally relevant metastable states with the goal of identifying optimal CVs that would structurally discriminate these states. Specific emphasis was on the automated detection and interpretive power of LDA-based CVs. We found that LDA methods, which included a novel gradient descent-based multiclass harmonic variant, termed GDHLDA, we developed here, outperform PCA in class separation, exhibiting remarkable consistency in extracting CVs critical for distinguishing metastable states. Furthermore, the identified CVs included features previously associated with conformational transitions in MFSD2A. Specifically, conformational shifts in transmembrane helix 7 and in residue Y294 on this helix emerged as critical features discriminating the metastable states in MFSD2A. This highlights the effectiveness of LDA-based approaches in automatically extracting from MD trajectories CVs of functional relevance that can be used to drive biased MD simulations to efficiently sample conformational transitions in the molecular system.

AI in Measurement Science

Measurement of biological systems containing biomolecules and bioparticles is a key task in the fields of analytical chemistry, biology, and medicine. Driven by the complex nature of biological systems and unprecedented amounts of measurement data, artificial intelligence (AI) in measurement science has rapidly advanced from the use of silicon-based machine learning (ML) for data mining to the development of molecular computing with improved sensitivity and accuracy. This review presents an overview of fundamental ML methodologies and discusses their applications in disease diagnostics, biomarker discovery, and imaging analysis. We next provide the working principles of molecular computing using logic gates and arithmetical devices, which can be employed for in situ detection, computation, and signal transduction for biological systems. This review concludes by summarizing the strengths and limitations of AI-involved biological measurement in fundamental and applied research.

Evaluating machine learning techniques for archaeological lithic sourcing: a case study of flint in Britain

It is 50 years since Sieveking et al. published their pioneering research in Nature on the geochemical analysis of artefacts from Neolithic flint mines in southern Britain. In the decades since, geochemical techniques to source stone artefacts have flourished globally, with a renaissance in recent years from new instrumentation, data analysis, and machine learning techniques. Despite the interest over these latter approaches, there has been variation in the quality with which these methods have been applied. Using the case study of flint artefacts and geological samples from England, we present a robust and objective evaluation of three popular techniques, Random Forest, K-Nearest-Neighbour, and Support Vector Machines, and present a pipeline for their appropriate use. When evaluated correctly, the results establish high model classification performance, with Random Forest leading with an average accuracy of 85% (measured through F1 Scores), and with Support Vector Machines following closely. The methodology developed in this paper demonstrates the potential to significantly improve on previous approaches, particularly in removing bias, and providing greater means of evaluation than previously utilised.

Machine learning approaches reveal subtle differences in breathing and sleep fragmentation in Phox2b-derived astrocytes ablated mice

Modern neurophysiology research requires the interrogation of high-dimensionality data sets. Machine learning and artificial intelligence (ML/AI) workflows have permeated into nearly all aspects of daily life in the developed world but have not been implemented routinely in neurophysiological analyses. The power of these workflows includes the speed at which they can be deployed, their availability of open-source programming languages, and the objectivity permitted in their data analysis. We used classification-based algorithms, including random forest, gradient boosted machines, support vector machines, and neural networks, to test the hypothesis that the animal genotypes could be separated into their genotype based on interpretation of neurophysiological recordings. We then interrogate the models to identify what were the major features utilized by the algorithms to designate genotype classification. By using raw EEG and respiratory plethysmography data, we were able to predict which recordings came from genotype class with accuracies that were significantly improved relative to the no information rate, although EEG analyses showed more overlap between groups than respiratory plethysmography. In comparison, conventional methods where single features between animal classes were analyzed, differences between the genotypes tested using baseline neurophysiology measurements showed no statistical difference. However, ML/AI workflows successfully were capable of providing successful classification, indicating that interactions between features were different in these genotypes. ML/AI workflows provide new methodologies to interrogate neurophysiology data. However, their implementation must be done with care so as to provide high rigor and reproducibility between laboratories. We provide a series of recommendations on how to report the utilization of ML/AI workflows for the neurophysiology community.NEW & NOTEWORTHY ML/AI classification workflows are capable of providing insight into differences between genotypes for neurophysiology research. Analytical techniques utilized in the neurophysiology community can be augmented by implementing ML/AI workflows. Random forest is a robust classification algorithm for respiratory plethysmography data. Utilization of ML/AI workflows in neurophysiology research requires heightened transparency and improved community research standards.

Evaluation of Forensic Data Using Logistic Regression-Based Classification Methods and an R Shiny Implementation

We demonstrate the use of classification methods that are well-suited for forensic toxicology applications. The methods are based on penalized logistic regression, can be employed when separation occurs in a two-class classification setting, and allow for the calculation of likelihood ratios. A case study of this framework is demonstrated on alcohol biomarker data for classifying chronic alcohol drinkers. The approach can be extended to applications in the fields of analytical and forensic chemistry, where it is a common feature to have a large number of biomarkers, and allows for flexibility in model assumptions such as multivariate normality. While some penalized regression methods have been introduced previously in forensic applications, our study is meant to encourage practitioners to use these powerful methods more widely. As such, based upon our proof-of-concept studies, we also introduce an R Shiny online tool with an intuitive interface able to perform several classification methods. We anticipate that this open-source and free-of-charge application will provide a powerful and dynamic tool to infer the LR value in case of classification tasks.

In vivo detection of murine glioblastoma through Raman and reflectance fiber-probe spectroscopies

Significance: Glioblastoma (GBM) is the most common and aggressive malignant brain tumor in adults. With a worldwide incidence rate of 2 to 3 per 100,000 people, it accounts for more than 60% of all brain cancers; currently, its 5-year survival rate is < 5 % . GBM treatment relies mainly on surgical resection. In this framework, multimodal optical spectroscopy could provide a fast and label-free tool for improving tumor detection and guiding the removal of diseased tissues. Aim: Discriminating healthy brain from GBM tissues in an animal model through the combination of Raman and reflectance spectroscopies. Approach: EGFP-GL261 cells were injected into the brains of eight laboratory mice for inducing murine GBM in these animals. A multimodal optical fiber probe combining fluorescence, Raman, and reflectance spectroscopy was used to localize in vivo healthy and tumor brain areas and to collect their spectral information. Results: Tumor areas were localized through the detection of EGFP fluorescence emission. Then, Raman and reflectance spectra were collected from healthy and tumor tissues, and later analyzed through principal component analysis and linear discriminant analysis in order to develop a classification algorithm. Raman and reflectance spectra resulted in 92% and 93% classification accuracy, respectively. Combining together these techniques allowed improving the discrimination between healthy and tumor tissues up to 97%. Conclusions: These preliminary results demonstrate the potential of multimodal fiber-probe spectroscopy for in vivo label-free detection and delineation of brain tumors, and thus represent an additional, encouraging step toward clinical translation and deployment of fiber-probe spectroscopy.

Abnormal Phase Coupling in Parkinson's Disease and Normalization Effects of Subthreshold Vestibular Stimulation

The human brain is a highly dynamic structure requiring dynamic coordination between different neural systems to perform numerous cognitive and behavioral tasks. Emerging perspectives on basal ganglia (BG) and thalamic functions have highlighted their role in facilitating and mediating information transmission among cortical regions. Thus, changes in BG and thalamic structures can induce aberrant modulation of cortico-cortical interactions. Recent work in deep brain stimulation (DBS) has demonstrated that externally applied electrical current to BG structures can have multiple downstream effects in large-scale brain networks. In this work, we identified EEG-based altered resting-state cortical functional connectivity in Parkinson's disease (PD) and examined effects of dopaminergic medication and electrical vestibular stimulation (EVS), a non-invasive brain stimulation (NIBS) technique capable of stimulating the BG and thalamus through vestibular pathways. Resting EEG was collected from 16 PD subjects and 18 age-matched, healthy controls (HC) in four conditions: sham (no stimulation), EVS1 (4-8 Hz multisine), EVS2 (50-100 Hz multisine) and EVS3 (100-150 Hz multisine). The mean, variability, and entropy were extracted from time-varying phase locking value (PLV), a non-linear measure of pairwise functional connectivity, to probe abnormal cortical couplings in the PD subjects. We found the mean PLV of Cz and C3 electrodes were important for discrimination between PD and HC subjects. In addition, the PD subjects exhibited lower variability and entropy of PLV (mostly in theta and alpha bands) compared to the controls, which were correlated with their clinical characteristics. While levodopa medication was effective in normalizing the mean PLV only, all EVS stimuli normalized the mean, variability and entropy of PLV in the PD subject, with the exact extent and duration of improvement a function of stimulus type. These findings provide evidence demonstrating both low- and high-frequency EVS exert widespread influences on cortico-cortical connectivity, likely via subcortical activation. The improvement observed in PD in a stimulus-dependent manner suggests that EVS with optimized parameters may provide a new non-invasive means for neuromodulation of functional brain networks.

Using positive matrix factorization to investigate microbial dehalogenation of chlorinated benzenes in groundwater at a historically contaminated site

Chlorinated benzenes are common groundwater contaminants in the United States, so demonstrating whether they undergo degradation in the subsurface is important in determining the best remedy for this contamination. The purpose of this work was to use a new data mining approach to investigate chlorinated benzene degradation pathways in the subsurface. Positive Matrix Factorization (PMF) was used to analyze long-term measurements of chlorinated benzene concentrations in groundwater at a contaminated site in New Jersey. A dataset containing 597 groundwater samples and 5 chlorinated benzenes and benzene collected from 144 wells over 20 years was investigated using PMF2 software. Despite the heterogeneity of this dataset, PMF analysis revealed patterns indicative of microbial dechlorination in the groundwater and provided insight about where dechlorination is occurring, to what extent, and under which geochemical conditions. PMF resolved a factor indicative of a source of 1,2,4-trichlorobenzene and 1,2-dichlorobenzene and two factors representing stages of dechlorination, one more advanced than the other. The PMF results indicated that virtually all of the 1,2-dichlorobenzene at the site arises from its use onsite, not from the dechlorination of trichlorobenzenes. Factors were further interpreted using ancillary data such as geochemical indicators and field parameters also measured in the samples. Analysis suggested that the partial and advanced dechlorination signals occur under different subsurface physical conditions. The results provided field validation of the current understanding of anaerobic dechlorination of chlorinated benzenes in the subsurface developed from laboratory studies. PMF is thereby shown to be a useful tool for investigating chlorinated benzene dechlorination.

CuS–Pt(iv)–PEG–FA nanoparticles for targeted photothermal and chemotherapy

CuS–Pt(iv) nanoparticles exhibited high in vitro and in vivo anti-tumor efficiency, which was caused by the integrated Pt drug-induced chemotherapy and CuS nanoparticle-mediated photothermal therapy (PTT) upon irradiation with near infrared (NIR) light.