Optical Nanosensing of Lipid Accumulation due to Enzyme Inhibition in Live Cells

SERS-Active Printable Hydrogel for 3D Cell Culture and Imaging

Hydrogel-based three-dimensional (3D) cell culture systems mimic the salient elements of extracellular matrices and promote native cell function. However, high-resolution 3D cell imaging that can provide biological information about multiple features of individual cells is yet to be realized. In this context, we incorporated plasmonic gold nanoparticles (AuNPs) into an alginate/gelatin hydrogel to produce surface-enhanced Raman spectroscopy (SERS)-active hydrogel inks for the 3D printing and culturing of Vero cells. Dense incorporation of AuNPs enables production of a printed 3D grid structure with 3D SERS performance, but with no measurable adverse effects on cell growth. Label-free SERS spectra were collected within a hydrogel, and a random forest binary classifier was developed to discriminate Vero cell signals from the hydrogel background with an accuracy of 87.5%. The results suggest that SERS signals from cellular components, such as proteins, lipids, and carbohydrates, account for this discrimination. We demonstrate visualization of cell shape, location, and density by combining predicted binary maps with peak feature intensity maps in 2D and 3D. SERS images with a resolution of ≈3 μm match well with the microscopy images and show clear increases in intensity with incubation time. We suggest that 3D SERS cell imaging is a promising means to examine the effect of external cell stimuli on cellular behavior for diagnostic purposes.

Opening the Random Forest Black Box of 1H NMR Metabolomics Data by the Exploitation of Surrogate Variables

The untargeted metabolomics analysis of biological samples with nuclear magnetic resonance (NMR) provides highly complex data containing various signals from different molecules. To use these data for classification, e.g., in the context of food authentication, machine learning methods are used. These methods are usually applied as a black box, which means that no information about the complex relationships between the variables and the outcome is obtained. In this study, we show that the random forest-based approach surrogate minimal depth (SMD) can be applied for a comprehensive analysis of class-specific differences by selecting relevant variables and analyzing their mutual impact on the classification model of different truffle species. SMD allows the assignment of variables from the same metabolites as well as the detection of interactions between different metabolites that can be attributed to known biological relationships.

In Situ Raman Study of Neurodegenerated Human Neuroblastoma Cells Exposed to Outer-Membrane Vesicles Isolated from Porphyromonas gingivalis

The aim of this study was to elucidate the chemistry of cellular degeneration in human neuroblastoma cells upon exposure to outer-membrane vesicles (OMVs) produced by Porphyromonas gingivalis (Pg) oral bacteria by monitoring their metabolomic evolution using in situ Raman spectroscopy. Pg-OMVs are a key factor in Alzheimer's disease (AD) pathogenesis, as they act as efficient vectors for the delivery of toxins promoting neuronal damage. However, the chemical mechanisms underlying the direct impact of Pg-OMVs on cell metabolites at the molecular scale still remain conspicuously unclear. A widely used in vitro model employing neuroblastoma SH-SY5Y cells (a sub-line of the SK-N-SH cell line) was spectroscopically analyzed in situ before and 6 h after Pg-OMV contamination. Concurrently, Raman characterizations were also performed on isolated Pg-OMVs, which included phosphorylated dihydroceramide (PDHC) lipids and lipopolysaccharide (LPS), the latter in turn being contaminated with a highly pathogenic class of cysteine proteases, a key factor in neuronal cell degradation. Raman characterizations located lipopolysaccharide fingerprints in the vesicle structure and unveiled so far unproved aspects of the chemistry behind protein degradation induced by Pg-OMV contamination of SH-SY5Y cells. The observed alterations of cells' Raman profiles were then discussed in view of key factors including the formation of amyloid β (Aβ) plaques and hyperphosphorylated Tau neurofibrillary tangles, and the formation of cholesterol agglomerates that exacerbate AD pathologies.

Exploitation of surrogate variables in random forests for unbiased analysis of mutual impact and importance of features

MOTIVATION

Random forest is a popular machine learning approach for the analysis of high-dimensional data because it is flexible and provides variable importance measures for the selection of relevant features. However, the complex relationships between the features are usually not considered for the selection and thus also neglected for the characterization of the analysed samples.

RESULTS

Here we propose two novel approaches that focus on the mutual impact of features in random forests. Mutual forest impact (MFI) is a relation parameter that evaluates the mutual association of the features to the outcome and, hence, goes beyond the analysis of correlation coefficients. Mutual impurity reduction (MIR) is an importance measure that combines this relation parameter with the importance of the individual features. MIR and MFI are implemented together with testing procedures that generate P-values for the selection of related and important features. Applications to one experimental and various simulated datasets and the comparison to other methods for feature selection and relation analysis show that MFI and MIR are very promising to shed light on the complex relationships between features and outcome. In addition, they are not affected by common biases, e.g. that features with many possible splits or high minor allele frequencies are preferred.

AVAILABILITY AND IMPLEMENTATION

The approaches are implemented in Version 0.3.3 of the R package RFSurrogates that is available at github.com/AGSeifert/RFSurrogates and the data are available at doi.org/10.25592/uhhfdm.12620.

Machine learning for nanoplasmonics

Plasmonic nanomaterials have outstanding optoelectronic properties potentially enabling the next generation of catalysts, sensors, lasers and photothermal devices. Owing to optical and electron techniques, modern nanoplasmonics research generates large datasets characterizing features across length scales. Furthermore, optimizing syntheses leading to specific nanostructures requires time-consuming multiparametric approaches. These complex datasets and trial-and-error practices make nanoplasmonics research ripe for the application of machine learning (ML) and advanced data processing methods. ML algorithms capture relationships between synthesis, structure and performance in a way that far exceeds conventional simulation and theory approaches, enabling effective performance optimization. For example, neural networks can tailor the nanostructure morphology to target desired properties, identify synthetic conditions and extract quantitative information from complex data. Here we discuss the nascent field of ML for nanoplasmonics, describe the opportunities and limitations of ML in nanoplasmonic research, and conclude that ML is potentially transformative, especially if the community curates and shares its big data.

SERS-Based Biosensors Combined with Machine Learning for Medical Application

Surface-enhanced Raman spectroscopy (SERS) has shown strength in non-invasive, rapid, trace analysis and has been used in many fields in medicine. Machine learning (ML) is an algorithm that can imitate human learning styles and structure existing content with the knowledge to effectively improve learning efficiency. Integrating SERS and ML can have a promising future in the medical field. In this review, we summarize the applications of SERS combined with ML in recent years, such as the recognition of biological molecules, rapid diagnosis of diseases, developing of new immunoassay techniques, and enhancing SERS capabilities in semi-quantitative measurements. Ultimately, the possible opportunities and challenges of combining SERS with ML are addressed.

Super-Long SERS Active Single Silver Nanowires for Molecular Imaging in 2D and 3D Cell Culture Models

Establishing a systematic molecular information analysis strategy for cell culture models is of great significance for drug development and tissue engineering technologies. Here, we fabricated single silver nanowires with high surface-enhanced Raman scattering activity to extract SERS spectra in situ from two-dimensional (2D) and three-dimensional (3D) cell culture models. The silver nanowires were super long, flexible and thin enough to penetrate through multiple cells. A single silver nanowire was used in combination with a four-dimensional microcontroller as a cell endoscope for spectrally analyzing the components in cell culture models. Then, we adopted a machine learning algorithm to analyze the obtained spectra. Our results show that the abundance of proteins differs significantly between the 2D and 3D models, and that nucleic acid-rich and protein-rich regions can be distinguished with satisfactory accuracy.

Advances in Nanomaterial-Based Biosensors for Determination of Glycated Hemoglobin

Diabetes is a major public health burden whose prevalence has been steadily increasing over the past decades. Glycated hemoglobin (HbA1c) is currently the gold standard for diagnostics and monitoring glycemic control in diabetes patients. HbA1c biosensors are often considered to be cost-effective alternatives for smaller testing laboratories or clinics unable to access other reference methods. Many of these sensors deploy nanomaterials as recognition elements, detection labels, and/or transducers for achieving sensitive and selective detection of HbA1c. Nanomaterials have emerged as important sensor components due to their excellent optical and electrical properties, tunable morphologies, and easy integration into multiple sensing platforms. In this review, we discuss the advantages of using nanomaterials to construct HbA1c sensors and various sensing strategies for HbA1c measurements. Key gaps between the current technologies with what is needed moving forward are also summarized.

Cancer Stem Cell Derived Extracellular Vesicles with Self-Functionalized 3D Nanosensor for Real-Time Cancer Diagnosis: Eliminating the Roadblocks in Liquid Biopsy

Liquid biopsy for determining the presence of cancer and the underlying tissue of origin is crucial to overcome the limitations of existing tissue biopsy and imaging-based techniques by capturing critical information from the dynamic tumor heterogeneity. A newly emerging liquid biopsy with extracellular vesicles (EVs) is gaining momentum, but its clinical relevance is in question due to the biological and technical challenges posed by existing technologies. The biological barriers of existing technologies include the inability to generate fundamental details of molecular structure, chemical composition as well as functional variations in EVs by gathering simultaneous information on multiple intra-EV molecules, unavailability of holistic qualitative analysis, in addition to the inability to identify tissue of origin. Technological barriers include reliance on EV isolation with a few labeled biomarkers, resulting in the inability to generate comprehensive information on the disease. A more favorable approach would be to generate holistic information on the disease without the use of labels. Such a marker-free diagnosis is impossible with the existing liquid biopsy due to the unavailability clinically validated cancer stem cells (CSC)-specific markers and dependence of existing technologies on EV isolation, undermining the clinical relevance of EV-based liquid biopsy. Here, CSC EVs were employed as an independent liquid biopsy modality. We hypothesize that tracking the signals of CSCs in peripheral blood with CSC EVs will provide a reliable solution for accurate cancer diagnosis, as CSC are the originators of tumor contributing to tumor heterogeneity. We report nanoengineered 3D sensors of extremely small nano-scaled probes self-functionalized for SERS, enabling integrative molecular and functional profiling of otherwise undetectable CSC EVs. A substantially enhanced SERS and ultralow limit of detection (10 EVs per 10 μL) were achieved. This was attributed to the efficient probe-EV interaction due to the 3D networks of nanoprobes, ensuring simultaneous detection of multiple EV signals. We experimentally demonstrate the crucial role of CSC EVs in cancer diagnosis. We then completed a pilot validation of this modality for cancer detection as well as for identification of the tissue of origin. An artificial neural network distinguished cancer from noncancer with 100% sensitivity and 100% specificity for three hard to detect cancers (breast, lung, and colorectal cancer). Binary classification to distinguish one tissue of origin against all other achieved 100% accuracy, while simultaneous identification of all three tissues of origin with multiclass classification achieved up to 79% accuracy. This noninvasive tool may complement existing cancer diagnostics, treatment monitoring as well as longitudinal disease monitoring by validation with a large cohort of clinical samples.

Structure and Interaction of Ceramide-Containing Liposomes with Gold Nanoparticles as Characterized by SERS and Cryo-EM

Due to the great potential of surface-enhanced Raman scattering (SERS) as local vibrational probe of lipid-nanostructure interaction in lipid bilayers, it is important to characterize these interactions in detail. The interpretation of SERS data of lipids in living cells requires an understanding of how the molecules interact with gold nanostructures and how intermolecular interactions influence the proximity and contact between lipids and nanoparticles. Ceramide, a sphingolipid that acts as important structural component and regulator of biological function, therefore of interest to probing, lacks a phosphocholine head group that is common to many lipids used in liposome models. SERS spectra of liposomes of a mixture of ceramide, phosphatidic acid, and phosphatidylcholine, as well as of pure ceramide and of the phospholipid mixture are reported. Distinct groups of SERS spectra represent varied contributions of the choline, sphingosine, and phosphate head groups and the structures of the acyl chains. Spectral bands related to the state of order of the membrane and moreover to the amide function of the sphingosine head groups indicate that the gold nanoparticles interact with molecules involved in different intermolecular relations. While cryogenic electron microscopy shows the formation of bilayer liposomes in all preparations, pure ceramide was found to also form supramolecular, concentric stacked and densely packed lamellar, nonliposomal structures. That the formation of such supramolecular assemblies supports the intermolecular interactions of ceramide is indicated by the SERS data. The unique spectral features that are assigned to the ceramide-containing lipid model systems here enable an identification of these molecules in biological systems and allow us to obtain information on their structure and interaction by SERS.

Surface enhanced Raman scattering for probing cellular biochemistry

Surface enhanced Raman scattering (SERS) from biomolecules in living cells enables the sensitive, but also very selective, probing of their biochemical composition. This minireview discusses the developments of SERS probing in cells over the past years from the proof-of-principle to observe a biochemical status to the characterization of molecule-nanostructure and molecule-molecule interactions and cellular processes that involve a wide variety of biomolecules and cellular compartments. Progress in applying SERS as a bioanalytical tool in living cells, to gain a better understanding of cellular physiology and to harness the selectivity of SERS, has been achieved by a combination of live cell SERS with several different approaches. They range from organelle targeting, spectroscopy of relevant molecular models, and the optimization of plasmonic nanostructures to the application of machine learning and help us to unify the information from defined biomolecules and from the cell as an extremely complex system.

Immunosuppressive calcineurin inhibitor cyclosporine A induces proapoptotic endoplasmic reticulum stress in renal tubular cells

Current immunosuppressive strategies in organ transplantation rely on calcineurin inhibitors cyclosporine A (CsA) or tacrolimus (Tac). Both drugs are nephrotoxic, but CsA has been associated with greater renal damage than Tac. CsA inhibits calcineurin by forming complexes with cyclophilins, whose chaperone function is essential for proteostasis. We hypothesized that stronger toxicity of CsA may be related to suppression of cyclophilins with ensuing endoplasmic reticulum (ER) stress and unfolded protein response (UPR) in kidney epithelia. Effects of CsA and Tac (10 µM for 6 h each) were compared in cultured human embryonic kidney 293 (HEK 293) cells, primary human renal proximal tubule (PT) cells, freshly isolated rat PTs, and knockout HEK 293 cell lines lacking the critical ER stress sensors, protein kinase RNA-like ER kinase or activating transcription factor 6 (ATF6). UPR was evaluated by detection of its key components. Compared with Tac treatment, CsA induced significantly stronger UPR in native cultured cells and isolated PTs. Evaluation of proapoptotic and antiapoptotic markers suggested an enhanced apoptotic rate in CsA-treated cells compared with Tac-treated cells as well. Similar to CsA treatment, knockdown of cyclophilin A or B by siRNA caused proapoptotic UPR, whereas application of the chemical chaperones tauroursodeoxycholic acid or 4-phenylbutyric acid alleviated CsA-induced UPR. Deletion of protein kinase RNA-like ER kinase or ATF6 blunted CsA-induced UPR as well. In summary, inhibition of cyclophilin chaperone function with ensuing ER stress and proapoptotic UPR aggravates CsA toxicity, whereas pharmacological modulation of UPR bears potential to alleviate renal side effects of CsA.

Opening the Random Forest Black Box of the Metabolome by the Application of Surrogate Minimal Depth

For the untargeted analysis of the metabolome of biological samples with liquid chromatography–mass spectrometry (LC-MS), high-dimensional data sets containing many different metabolites are obtained. Since the utilization of these complex data is challenging, different machine learning approaches have been developed. Those methods are usually applied as black box classification tools, and detailed information about class differences that result from the complex interplay of the metabolites are not obtained. Here, we demonstrate that this information is accessible by the application of random forest (RF) approaches and especially by surrogate minimal depth (SMD) that is applied to metabolomics data for the first time. We show this by the selection of important features and the evaluation of their mutual impact on the multi-level classification of white asparagus regarding provenance and biological identity. SMD enables the identification of multiple features from the same metabolites and reveals meaningful biological relations, proving its high potential for the comprehensive utilization of high-dimensional metabolomics data.

Influence of Nuclear Localization Sequences on the Intracellular Fate of Gold Nanoparticles

Directing nanoparticles to the nucleus by attachment of nuclear localization sequences (NLS) is an aim in many applications. Gold nanoparticles modified with two different NLS were studied while crossing barriers of intact cells, including uptake, endosomal escape, and nuclear translocation. By imaging of the nanoparticles and by characterization of their molecular interactions with surface-enhanced Raman scattering (SERS), it is shown that nuclear translocation strongly depends on the particular incubation conditions. After an 1 h of incubation followed by a 24 h chase time, 14 nm gold particles carrying an adenoviral NLS are localized in endosomes, in the cytoplasm, and in the nucleus of fibroblast cells. In contrast, the cells display no nanoparticles in the cytoplasm or nucleus when continuously incubated with the nanoparticles for 24 h. The ultrastructural and spectroscopic data indicate different processing of NLS-functionalized particles in endosomes compared to unmodified particles. NLS-functionalized nanoparticles form larger intraendosomal aggregates than unmodified gold nanoparticles. SERS spectra of cells with NLS-functionalized gold nanoparticles contain bands assigned to DNA and were clearly different from those with unmodified gold nanoparticles. The different processing in the presence of an NLS is influenced by a continuous exposure of the cells to nanoparticles and an ongoing nanoparticle uptake. This is supported by mass-spectrometry-based quantification that indicates enhanced uptake of NLS-functionalized nanoparticles compared to unmodified particles under the same conditions. The results contribute to the optimization of nanoparticle analysis in cells in a variety of applications, e.g., in theranostics, biotechnology, and bioanalytics.

Molecular Structure and Interactions of Lipids in the Outer Membrane of Living Cells Based on Surface-Enhanced Raman Scattering and Liposome Models

The distribution and interaction of lipids determine the structure and function of the cellular membrane. Surface-enhanced Raman scattering (SERS) is used for selective molecular probing of the cell membrane of living fibroblast cells grown adherently on gold nanoisland substrates across their whole contact areas with the substrate, enabling mapping of the membrane's composition and interaction. From the SERS data, the localization and distribution of different lipids and their interactions, together with proteins in the outer cell membrane, are inferred. Interpretation of the spectra is mainly supported by comparison with the spectra of model liposomes composed of phosphatidylcholine, sphingomyelin, and cholesterol obtained on the same gold substrate. The interaction of the liposomes with the substrate differs from that with gold nanoparticles. The SERS maps indicate colocalization of ordered lipid domains with cholesterol in the living cells. They support the observation of ordered membrane regions of micrometer dimensions in the outer leaflet of the cell membrane that are rich in sphingomyelin. Moreover, the spectra of the living cells contain bands from the groups of the lipid heads, phosphate, choline, and ethanolamine, combined with those from membrane proteins, as indicated by signals assigned to prenyl attachment. Elucidating the composition and structure of lipid membranes in living cells can find application in many fields of research.

Probing the Intracellular Bio-Nano Interface in Different Cell Lines with Gold Nanostars

Gold nanostars are a versatile plasmonic nanomaterial with many applications in bioanalysis. Their interactions with animal cells of three different cell lines are studied here at the molecular and ultrastructural level at an early stage of endolysosomal processing. Using the gold nanostars themselves as substrate for surface-enhanced Raman scattering, their protein corona and the molecules in the endolysosomal environment were characterized. Localization, morphology, and size of the nanostar aggregates in the endolysosomal compartment of the cells were probed by cryo soft-X-ray nanotomography. The processing of the nanostars by macrophages of cell line J774 differed greatly from that in the fibroblast cell line 3T3 and in the epithelial cell line HCT-116, and the structure and composition of the biomolecular corona was found to resemble that of spherical gold nanoparticles in the same cells. Data obtained with gold nanostars of varied morphology indicate that the biomolecular interactions at the surface in vivo are influenced by the spike length, with increased interaction with hydrophobic groups of proteins and lipids for longer spike lengths, and independent of the cell line. The results will support optimized nanostar synthesis and delivery for sensing, imaging, and theranostics.

Intracellular optical probing with gold nanostars

Gold nanostars are important nanoscopic tools in biophotonics and theranostics. To understand the fate of such nanostructures in the endolysosomal system of living cells as an important processing route in biotechnological approaches, un-labelled, non-targeted gold nanostars synthesized using HEPES buffer were studied in two cell lines. The uptake of the gold nanostructures leads to cell line-dependent intra-endolysosomal agglomeration, which results in a greater enhancement of the local optical fields than those around individual nanostars and near aggregates of spherical gold nanoparticles of the same size. As demonstrated by non-resonant surface-enhanced Raman scattering (SERS) spectra in the presence and absence of aggregation, the spectroscopic signals of molecules are of very similar strength over a wide range of concentrations, which is ideal for label-free vibrational characterization of cells and other complex environments. In 3T3 and HCT-116 cells, SERS data were analyzed together with the properties of the intracellular nanostar agglomerates. Vibrational spectra indicate that the processing of nanostars by cells and their interaction with the surrounding endolysosomal compartment is connected to their morphological properties through differences in the structure and interactions in their intracellular protein corona. Specifically, different intracellular processing was found to result from a different extent of hydrophobic interactions at the pristine gold surface, which varies for nanostars of different spike lengths. The sensitive optical monitoring of surroundings of nanostars and their intracellular processing makes them a very useful tool for optical bionanosensing and therapy.

Predicting nanotoxicity by an integrated machine learning and metabolomics approach

Predicting the biological responses to engineered nanoparticles (ENPs) is critical to their environmental health assessment. The disturbances of metabolic pathways reflect the global profile of biological responses to ENPs but are difficult to predict due to the highly heterogeneous data from complicated biological systems and various ENP properties. Herein, integrating multiple machine learning models and metabolomics enabled accurate prediction of the disturbance of metabolic pathways induced by 33 ENPs. Screening nine typical properties of ENPs identified type and size as the top features determining the effects on metabolic pathways. Similarity network analysis and decision tree models overcame the highly heterogeneous data sources to visualize and judge the occurrence of metabolic pathways depending on the sorting priority features. The model accuracy was verified by animal experiments and reached 75%-100%, even for the prediction of ENPs outside of databases. The models also predicted metabolic pathway-related histopathology. This work provides an approach for the quick assessment of environmental health risks induced by known and unknown ENPs.

Relating the composition and interface interactions in the hard corona of gold nanoparticles to the induced response mechanisms in living cells

Understanding the formation of the intracellular protein corona of nanoparticles is essential for a wide range of bio- and nanomedical applications. The innermost layer of the protein corona, the hard corona, directly interacts with the nanoparticle surface, and by shielding the surface, it has a deterministic effect on the intracellular processing of the nanoparticle. Here, we combine a direct qualitative analysis of the hard corona composition of gold nanoparticles with a detailed structural characterization of the molecules in their interaction with the nanoparticle surface and relate both to the effects they have on the ultrastructure of living cells and the processing of the gold nanoparticles. Cells from the cell lines HCT-116 and A549 were incubated with 30 nm citrate-stabilized gold nanoparticles and with their aggregates in different culture media. The combined results of mass spectrometry based proteomics, cryo soft X-ray nanotomography and surface-enhanced Raman scattering experiments together revealed different uptake mechanisms in the two cell lines and distinct levels of induced cellular stress when incubation conditions were varied. The data indicate that the different incubation conditions lead to changes in the nanoparticle processing via different protein-nanoparticle interfacial interactions. Specifically, they suggest that the protein-nanoparticle surface interactions depend mainly on the surface properties of the gold nanoparticles, that is, the ζ-potential and the resulting changes in the hydrophilicity of the nanoparticle surface, and are largely independent of the cell line, the uptake mechanism and intracellular processing, or the extent of the induced cellular stress.

Early-Stage Lung Cancer Diagnosis by Deep Learning-Based Spectroscopic Analysis of Circulating Exosomes

Lung cancer has a high mortality rate, but an early diagnosis can contribute to a favorable prognosis. A liquid biopsy that captures and detects tumor-related biomarkers in body fluids has great potential for early-stage diagnosis. Exosomes, nanosized extracellular vesicles found in blood, have been proposed as promising biomarkers for liquid biopsy. Here, we demonstrate an accurate diagnosis of early-stage lung cancer, using deep learning-based surface-enhanced Raman spectroscopy (SERS) of the exosomes. Our approach was to explore the features of cell exosomes through deep learning and figure out the similarity in human plasma exosomes, without learning insufficient human data. The deep learning model was trained with SERS signals of exosomes derived from normal and lung cancer cell lines and could classify them with an accuracy of 95%. In 43 patients, including stage I and II cancer patients, the deep learning model predicted that plasma exosomes of 90.7% patients had higher similarity to lung cancer cell exosomes than the average of the healthy controls. Such similarity was proportional to the progression of cancer. Notably, the model predicted lung cancer with an area under the curve (AUC) of 0.912 for the whole cohort and stage I patients with an AUC of 0.910. These results suggest the great potential of the combination of exosome analysis and deep learning as a method for early-stage liquid biopsy of lung cancer.

Application of random forest based approaches to surface-enhanced Raman scattering data

Surface-enhanced Raman scattering (SERS) is a valuable analytical technique for the analysis of biological samples. However, due to the nature of SERS it is often challenging to exploit the generated data to obtain the desired information when no reporter or label molecules are used. Here, the suitability of random forest based approaches is evaluated using SERS data generated by a simulation framework that is also presented. More specifically, it is demonstrated that important SERS signals can be identified, the relevance of predefined spectral groups can be evaluated, and the relations of different SERS signals can be analyzed. It is shown that for the selection of important SERS signals Boruta and surrogate minimal depth (SMD) and for the analysis of spectral groups the competing method Learner of Functional Enrichment (LeFE) should be applied. In general, this investigation demonstrates that the combination of random forest approaches and SERS data is very promising for sophisticated analysis of complex biological samples.