Metabolic Fingerprinting Links Oncogenic PIK3CA with Enhanced Arachidonic Acid-Derived Eicosanoids.

Oncogenic transformation is associated with profound changes in cellular metabolism, but whether tracking these can improve disease stratification or influence therapy decision-making is largely unknown. Using the iKnife to sample the aerosol of cauterized specimens, we demonstrate a new mode of real-time diagnosis, coupling metabolic phenotype to mutant PIK3CA genotype. Oncogenic PIK3CA results in an increase in arachidonic acid and a concomitant overproduction of eicosanoids, acting to promote cell proliferation beyond a cell-autonomous manner. Mechanistically, mutant PIK3CA drives a multimodal signaling network involving mTORC2-PKCζ-mediated activation of the calcium-dependent phospholipase A2 (cPLA2). Notably, inhibiting cPLA2 synergizes with fatty acid-free diet to restore immunogenicity and selectively reduce mutant PIK3CA-induced tumorigenicity. Besides highlighting the potential for metabolic phenotyping in stratified medicine, this study reveals an important role for activated PI3K signaling in regulating arachidonic acid metabolism, uncovering a targetable metabolic vulnerability that largely depends on dietary fat restriction. VIDEO ABSTRACT.

A targeted LC-MRM3 proteomic approach for the diagnosis of SARS-CoV-2 infection in nasopharyngeal swabs.

Since its first appearance, SARS-CoV-2 quickly spread around the world and the lack of adequate PCR testing capacities, especially during the early pandemic, led the scientific community to explore new approaches such as mass spectrometry (MS). We developed a proteomics workflow to target several tryptic peptides of the nucleocapsid protein (NCAP). A highly selective multiple reaction monitoring MRM3 strategy provided a sensitivity increase in comparison to conventional MRM acquisition. Our MRM3 approach was first tested on an Amsterdam public health cohort (alpha-variant, 760 participants) detecting viral NCAP peptides from nasopharyngeal swabs samples presenting a cycle threshold (Ct) value down to 35 with sensitivity and specificity of 94.2% and 100.0%, without immuno-purification. A second iteration of the MS-diagnostic test, able to analyze more than 400 samples per day, was clinically validated on a Leiden-Rijswijk public health cohort (delta-variant, 2536 participants) achieving 99.9% specificity and 93.1% sensitivity for patients with Ct-values up to 35. In this manuscript, we also developed and brought the first proof of the concept of viral variant monitoring in a complex matrix using targeted mass spectrometry.

Mass spectrometry imaging highlights dynamic patterns of lipid co-expression with Aß plaques in mouse and human brains.

Lipids play crucial roles in the susceptibility and brain cellular responses to Alzheimer's disease (AD) and are increasingly considered potential soluble biomarkers in cerebrospinal fluid (CSF) and plasma. To delineate the pathological correlations of distinct lipid species, we conducted a comprehensive characterization of both spatially localized and global differences in brain lipid composition in AppNL-G-F mice with spatial and bulk mass spectrometry lipidomic profiling, using human amyloid-expressing (h-Aß) and WT mouse brains controls. We observed age-dependent increases in lysophospholipids, bis(monoacylglycerol) phosphates, and phosphatidylglycerols around Aß plaques in AppNL-G-F mice. Immunohistology-based co-localization identified associations between focal pro-inflammatory lipids, glial activation, and autophagic flux disruption. Likewise, in human donors with varying Braak stages, similar studies of cortical sections revealed co-expression of lysophospholipids and ceramides around Aß plaques in AD (Braak stage V/VI) but not in earlier Braak stage controls. Our findings in mice provide evidence of temporally and spatially heterogeneous differences in lipid composition as local and global Aß-related pathologies evolve. Observing similar lipidomic changes associated with pathological Aß plaques in human AD tissue provides a foundation for understanding differences in CSF lipids with reported clinical stage or disease severity.

Clinicopathological Predictors of Positive Resection Margins in Breast-Conserving Surgery.

BACKGROUND: Ductal carcinoma in situ (DCIS) is associated with risk of positive resection margins following breast-conserving surgery (BCS) and subsequent reoperation. Prior reports grossly underestimate the risk of margin positivity with IBC containing a DCIS component (IBC + DCIS) due to patient-level rather than margin-level analysis. OBJECTIVE: The aim of this study was to delineate the relative risk of IBC + DCIS compared with pure IBC (without a DCIS component) on margin positivity through detailed margin-level interrogation. METHODS: A single institution, retrospective, observational cohort study was conducted in which pathology databases were evaluated to identify patients who underwent BCS over 5 years (2014-2019). Margin-level interrogation included granular detail into the extent, pathological subtype and grade of disease at each resection margin. Predictors of a positive margin were computed using multivariate regression analysis. RESULTS: Clinicopathological details were examined from 5454 margins from 909 women. The relative risk of a positive margin with IBC + DCIS versus pure IBC was 8.76 (95% confidence interval [CI] 6.64-11.56) applying UK Association of Breast Surgery guidelines, and 8.44 (95% CI 6.57-10.84) applying the Society of Surgical Oncology/American Society for Radiation Oncology guidelines. Independent predictors of margin positivity included younger patient age (0.033, 95% CI 0.006-0.060), lower specimen weight (0.045, 95% CI 0.020-0.069), multifocality (0.256, 95% CI 0.137-0.376), lymphovascular invasion (0.138, 95% CI 0.068-0.208) and comedonecrosis (0.113, 95% CI 0.040-0.185). CONCLUSIONS: Compared with pure IBC, the relative risk of a positive margin with IBC + DCIS is approximately ninefold, significantly higher than prior estimates. This margin-level methodology is believed to represent the impact of DCIS more accurately on margin positivity in IBC.

Lipidomic Profiling of Colorectal Lesions for Real-Time Tissue Recognition and Risk-Stratification Using Rapid Evaporative Ionization Mass Spectrometry.

OBJECTIVE: Rapid evaporative ionization mass spectrometry (REIMS) is a metabolomic technique analyzing tissue metabolites, which can be applied intraoperatively in real-time. The objective of this study was to profile the lipid composition of colorectal tissues using REIMS, assessing its accuracy for real-time tissue recognition and risk-stratification. SUMMARY BACKGROUND DATA: Metabolic dysregulation is a hallmark feature of carcinogenesis; however, it remains unknown if this can be leveraged for real-time clinical applications in colorectal disease. METHODS: Patients undergoing colorectal resection were included, with carcinoma, adenoma and paired-normal mucosa sampled. Ex vivo analysis with REIMS was conducted using monopolar diathermy, with the aerosol aspirated into a Xevo G2S QToF mass spectrometer. Negatively charged ions over 600 to 1000 m/z were used for univariate and multivariate functions including linear discriminant analysis. RESULTS: A total of 161 patients were included, generating 1013 spectra. Unique lipidomic profiles exist for each tissue type, with REIMS differentiating samples of carcinoma, adenoma, and normal mucosa with 93.1% accuracy and 96.1% negative predictive value for carcinoma. Neoplasia (carcinoma or adenoma) could be predicted with 96.0% accuracy and 91.8% negative predictive value. Adenomas can be risk-stratified by grade of dysplasia with 93.5% accuracy, but not histological subtype. The structure of 61 lipid metabolites was identified, revealing that during colorectal carcinogenesis there is progressive increase in relative abundance of phosphatidylglycerols, sphingomyelins, and mono-unsaturated fatty acid-containing phospholipids. CONCLUSIONS: The colorectal lipidome can be sampled by REIMS and leveraged for accurate real-time tissue recognition, in addition to riskstratification of colorectal adenomas. Unique lipidomic features associated with carcinogenesis are described.

The postprandial secretion of peptide YY1-36 and 3-36 in obesity is differentially increased after gastric bypass versus sleeve gastrectomy.

OBJECTIVES: Peptide tyrosine tyrosine (PYY) exists as two species, PYY1-36 and PYY3-36 , with distinct effects on insulin secretion and appetite regulation. The detailed effects of bariatric surgery on PYY1-36 and PYY3-36 secretion are not known as previous studies have used nonspecific immunoassays to measure total PYY. Our objective was to characterize the effect of sleeve gastrectomy (SG) and Roux-en-Y gastric bypass (RYGB) on fasting and postprandial PYY1-36 and PYY3-36 secretion using a newly developed liquid chromatography-tandem mass spectrometry (LC-MS/MS) assay. DESIGN AND SUBJECTS: Observational study in 10 healthy nonobese volunteers and 30 participants with obesity who underwent RYGB (n = 24) or SG (n = 6) at the Imperial Weight Centre [NCT01945840]. Participants were studied using a standardized mixed meal test (MMT) before and 1 year after surgery. The outcome measures were PYY1-36 and PYY3-36 concentrations. RESULTS: Presurgery, the fasting and postprandial levels of PYY1-36 and PYY3-36 were low, with minimal responses to the MMT, and these did not differ from healthy nonobese volunteers. The postprandial secretion of both PYY1-36 and PYY3-36 at 1 year was amplified after RYGB, but not SG, with the response being significantly higher in RYGB compared with SG. CONCLUSIONS: There appears to be no difference in PYY secretion between nonobese and obese volunteers at baseline. At 1 year after surgery, RYGB, but not SG, is associated with increased postprandial secretion of PYY1-36 and PYY3-36 , which may account for long-term differences in efficacy and adverse effects between the two types of surgery.

The intelligent knife (iKnife) and its intraoperative diagnostic advantage for the treatment of cervical disease.

Clearance of surgical margins in cervical cancer prevents the need for adjuvant chemoradiation and allows fertility preservation. In this study, we determined the capacity of the rapid evaporative ionization mass spectrometry (REIMS), also known as intelligent knife (iKnife), to discriminate between healthy, preinvasive, and invasive cervical tissue. Cervical tissue samples were collected from women with healthy, human papilloma virus (HPV) ± cervical intraepithelial neoplasia (CIN), or cervical cancer. A handheld diathermy device generated surgical aerosol, which was transferred into a mass spectrometer for subsequent chemical analysis. Combination of principal component and linear discriminant analysis and least absolute shrinkage and selection operator was employed to study the spectral differences between groups. Significance of discriminatory m/z features was tested using univariate statistics and tandem MS performed to elucidate the structure of the significant peaks allowing separation of the two classes. We analyzed 87 samples (normal = 16, HPV ± CIN = 50, cancer = 21 patients). The iKnife discriminated with 100% accuracy normal (100%) vs. HPV ± CIN (100%) vs. cancer (100%) when compared to histology as the gold standard. When comparing normal vs. cancer samples, the accuracy was 100% with a sensitivity of 100% (95% CI 83.9 to 100) and specificity 100% (79.4 to 100). Univariate analysis revealed significant MS peaks in the cancer-to-normal separation belonging to various classes of complex lipids. The iKnife discriminates healthy from premalignant and invasive cervical lesions with high accuracy and can improve oncological outcomes and fertility preservation of women treated surgically for cervical cancer. Larger in vivo research cohorts are required to validate these findings.

Mass recalibration for desorption electrospray ionization mass spectrometry imaging using endogenous reference ions.

BACKGROUND: Mass spectrometry imaging (MSI) data often consist of tens of thousands of mass spectra collected from a sample surface. During the time necessary to perform a single acquisition, it is likely that uncontrollable factors alter the validity of the initial mass calibration of the instrument, resulting in mass errors of magnitude significantly larger than their theoretical values. This phenomenon has a two-fold detrimental effect: (a) it reduces the ability to interpret the results based on the observed signals, (b) it can affect the quality of the observed signal spatial distributions. RESULTS: We present a post-acquisition computational method capable of reducing the observed mass drift by up to 60 ppm in biological samples, exploiting the presence of typical molecules with a known mass-to-charge ratio. The procedure, tested on time-of-flight and Orbitrap mass spectrometry analyzers interfaced to a desorption electrospray ionization (DESI) source, improves the molecular annotation quality and the spatial distributions of the detected ions. CONCLUSION: The presented method represents a robust and accurate tool for performing post-acquisition mass recalibration of DESI-MSI datasets and can help to increase the reliability of the molecular assignment and the data quality.

Atmospheric-Pressure Infrared Laser-Ablation Plasma-Postionization Mass Spectrometry Imaging of Formalin-Fixed Paraffin-Embedded (FFPE) and Fresh-Frozen Tissue Sections with No Sample Preparation.

Mass spectrometry imaging (MSI) encompasses a powerful suit of techniques which provide spatially resolved atomic and molecular information from almost any sample type. MSI is now widely used in preclinical research to provide insight into metabolic phenotypes of disease. Typically, fresh-frozen tissue preparations are considered optimal for biological MSI and other traditional preservation methods such as formalin fixation, alone or with paraffin embedding (FFPE), are considered less optimal or even incompatible. Due to the prevalence of FFPE tissue storage, particularly for rare and therefore high-value tissue samples, there is substantial motivation for optimizing MSI methods for analysis of FFPE tissue. Here, we present a novel modality, atmospheric-pressure infrared laser-ablation plasma postionization (AP-IR-LA-PPI), with the first proof-of-concept examples of MSI for FFPE and fresh-frozen tissues, with no post-sectioning sample preparation. We present ion images from FFPE and fresh tissues in positive and negative ion modes. Molecular annotations (via the Metaspace annotation engine) and on-tissue MS/MS provide additional confidence that the detected ions arise from a broad range of metabolite and lipid classes from both FFPE and fresh-frozen tissues.

1H NMR Signals from Urine Excreted Protein Are a Source of Bias in Probabilistic Quotient Normalization.

Normalization to account for variation in urinary dilution is crucial for interpretation of urine metabolic profiles. Probabilistic quotient normalization (PQN) is used routinely in metabolomics but is sensitive to systematic variation shared across a large proportion of the spectral profile (>50%). Where 1H nuclear magnetic resonance (NMR) spectroscopy is employed, the presence of urinary protein can elevate the spectral baseline and substantially impact the resulting profile. Using 1H NMR profile measurements of spot urine samples collected from hospitalized COVID-19 patients in the ISARIC 4C study, we determined that PQN coefficients are significantly correlated with observed protein levels (r2 = 0.423, p < 2.2 × 10-16). This correlation was significantly reduced (r2 = 0.163, p < 2.2 × 10-16) when using a computational method for suppression of macromolecular signals known as small molecule enhancement spectroscopy (SMolESY) for proteinic baseline removal prior to PQN. These results highlight proteinuria as a common yet overlooked source of bias in 1H NMR metabolic profiling studies which can be effectively mitigated using SMolESY or other macromolecular signal suppression methods before estimation of normalization coefficients.

High Resolution Ambient MS Imaging of Biological Samples by Desorption Electro-Flow Focussing Ionization.

In this study, we examine the suitability of desorption electro-flow focusing ionization (DEFFI) for mass spectrometry imaging (MSI) of biological tissue. We also compare the performance of desorption electrospray ionization (DESI) with and without the flow focusing setup. The main potential advantages of applying the flow focusing mechanism in DESI is its rotationally symmetric electrospray jet, higher intensity, more controllable parameters, and better portability due to the robustness of the sprayer. The parameters for DEFFI have therefore been thoroughly optimized, primarily for spatial resolution but also for intensity. Once the parameters have been optimized, DEFFI produces similar images to the existing DESI. MS images for mouse brain samples, acquired at a nominal pixel size of 50 µm, are comparable for both DESI setups, albeit the new sprayer design yields better sensitivity. Furthermore, the two methods are compared with regard to spectral intensity as well as the area of the desorbed crater on rhodamine-coated slides. Overall, the implementation of a flow focusing mechanism in DESI is shown to be highly suitable for imaging biological tissue and has potential to overcome some of the shortcomings experienced with the current geometrical design of DESI.

Method To Visualize the Intratumor Distribution and Impact of Gemcitabine in Pancreatic Ductal Adenocarcinoma by Multimodal Imaging.

Gemcitabine (dFdC) is a common treatment for pancreatic cancer; however, it is thought that treatment may fail because tumor stroma prevents drug distribution to tumor cells. Gemcitabine is a pro-drug with active metabolites generated intracellularly; therefore, visualizing the distribution of parent drug as well as its metabolites is important. A multimodal imaging approach was developed using spatially coregistered mass spectrometry imaging (MSI), imaging mass cytometry (IMC), multiplex immunofluorescence microscopy (mIF), and hematoxylin and eosin (H&E) staining to assess the local distribution and metabolism of gemcitabine in tumors from a genetically engineered mouse model of pancreatic cancer (KPC) allowing for comparisons between effects in the tumor tissue and its microenvironment. Mass spectrometry imaging (MSI) enabled the visualization of the distribution of gemcitabine (100 mg/kg), its phosphorylated metabolites dFdCMP, dFdCDP and dFdCTP, and the inactive metabolite dFdU. Distribution was compared to small-molecule ATR inhibitor AZD6738 (25 mg/kg), which was codosed. Gemcitabine metabolites showed heterogeneous distribution within the tumor, which was different from the parent compound. The highest abundance of dFdCMP, dFdCDP, and dFdCTP correlated with distribution of endogenous AMP, ADP, and ATP in viable tumor cell regions, showing that gemcitabine active metabolites are reaching the tumor cell compartment, while AZD6738 was located to nonviable tumor regions. The method revealed that the generation of active, phosphorylated dFdC metabolites as well as treatment-induced DNA damage primarily correlated with sites of high proliferation in KPC PDAC tumor tissue, rather than sites of high parent drug abundance.

peakPantheR, an R package for large-scale targeted extraction and integration of annotated metabolic features in LC-MS profiling datasets.

SUMMARY: Untargeted liquid chromatography-mass spectrometry (LC-MS) profiling assays are capable of measuring thousands of chemical compounds in a single sample, but unreliable feature extraction and metabolite identification remain considerable barriers to their interpretation and usefulness. peakPantheR (Peak Picking and ANnoTation of High-resolution Experiments in R) is an R package for the targeted extraction and integration of annotated features from LC-MS profiling experiments. It takes advantage of chromatographic and spectral databases and prior information of sample matrix composition to generate annotated and interpretable metabolic phenotypic datasets and power workflows for real-time data quality assessment. AVAILABILITY AND IMPLEMENTATION: peakPantheR is available via Bioconductor (https://bioconductor.org/packages/peakPantheR/). Documentation and worked examples are available at https://phenomecentre.github.io/peakPantheR.github.io/ and https://github.com/phenomecentre/metabotyping-dementia-urine. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Diagnostic Accuracy of Nipple Discharge Fluid Cytology: A Meta-Analysis and Systematic Review of the Literature.

BACKGROUND: Nipple discharge is the third most frequent complaint of women attending rapid diagnostic breast clinics. Nipple smear cytology remains the single most used diagnostic method for investigating fluid content. This study aimed to conduct a systematic review and meta-analysis of the diagnostic accuracy of nipple discharge fluid assessment. METHODS: The study incorporated searches for studies interrogating the diagnostic data of nipple discharge fluid cytology compared with the histopathology gold standard. Data from studies published from 1956 to 2019 were analyzed. The analysis included 8648 cytology samples of women with a presenting complaint of nipple discharge. Both hierarchical and bivariate models for diagnostic meta-analysis were used to attain overall pooled sensitivity and specificity. RESULTS: Of 837 studies retrieved, 45 fulfilled the criteria for inclusion. The diagnostic accuracy of the meta-analysis examining nipple discharge fluid had a sensitivity of 75 % (95 % confidence interval [CI], 0.74-0.77) and a specificity of 87 % (95 % CI, 0.86-0.87) for benign breast disease. For breast cancer, it had a sensitivity of 62 % (95 % CI, 0.53-0.71) and a specificity 71 % (95 % CI, 0.57-0.81). Furthermore, patients presenting with blood-stained discharge yielded an overall malignancy rate of 58 % (95 % CI, 0.54-0.60) with a positive predictive value (PPV) of 27 % (95 % CI, 0.17-0.36). CONCLUSIONS: Pooled data from studies encompassing nipple discharge fluid assessment suggest that nipple smear cytology is of limited diagnostic accuracy. The authors recommend that a tailored approach to diagnosis be required given the variable sensitivities of currently available tests.

Modality Agnostic Model for Spatial Resolution in Mass Spectrometry Imaging: Application to MALDI MSI Data.

Image resolution in mass spectrometry imaging (MSI) is governed by the sampling probe, the motion of the stage relative to the probe, and the noise inherent for the sample and instrumentation employed. A new image formation model accounting for these variables is presented here. The model shows that the size of the probe, stage velocity, and the rate at which the probe consumes material from the surface govern the amount of blur present in the image. However, the main limiting factor for resolution is the signal-to-noise ratio (SNR). To evaluate blurring and noise effects, a new computational method for measuring lateral resolution in MSI is proposed. A spectral decomposition of the observed image signal and noise is used to determine a resolution number. To evaluate this technique, a silver step edge was prepared. This device was imaged at different pixels sizes using matrix-assisted laser desorption/ionization mass spectrometry imaging (MALDI MSI). A modulation transfer function (MTF) and a noise power spectrum (NPS) were computed for each single-ion image, and resolution was defined as the point of intersection between the MTF and the NPS. Finally, the algorithm was also applied to a MALDI MSI tissue data set.

Validation of Ultrasonic Harmonic Scalpel for Real-Time Tissue Identification Using Rapid Evaporative Ionization Mass Spectrometry.

In this study, we integrate rapid evaporative ionization mass spectrometry (REIMS) with the Harmonic scalpel, an advanced laparoscopic surgical instrument that utilizes ultrasound energy to dissect and coagulate tissues. It provides unparalleled manipulation capability to surgeons and has superseded traditional electrosurgical tools particularly in abdominal surgery, but is yet to be validated with REIMS. The REIMS platform coupled with the Harmonic device was shown to produce tissue-specific lipid profiles through the analysis of porcine tissues in both negative and positive ionization modes. Comparison with other methods of electrosurgical dissection, such as monopolar electrosurgery and CO2 laser, showed spectral differences in the profile dependent on the energy device used. The Harmonic device demonstrated major spectral differences in the phospholipid region of m/z 600-1000 compared with the monopolar electrosurgical and CO2 laser-generated spectra. Within the Harmonic REIMS spectra, high intensities of diglycerides and triglycerides were observed. In contrast, monopolar electrosurgical and laser spectra demonstrated high abundances of glycerophospholipids. The Harmonic scalpel was able to differentiate between the liver, muscle, colon, and small intestine, demonstrating 100% diagnostic accuracy. The validation of the Harmonic device-mass spectrometry combination will allow the platform to be used safely and robustly for real-time in vivo surgical tissue identification in a variety of clinical applications.

Representing the Metabolome with High Fidelity: Range and Response as Quality Control Factors in LC-MS-Based Global Profiling.

Liquid chromatography-mass spectrometry (LC-MS) is a powerful and widely used technique for measuring the abundance of chemical species in living systems. Its sensitivity, analytical specificity, and direct applicability to biofluids and tissue extracts impart great promise for the discovery and mechanistic characterization of biomarker panels for disease detection, health monitoring, patient stratification, and treatment personalization. Global metabolic profiling applications yield complex data sets consisting of multiple feature measurements for each chemical species observed. While this multiplicity can be useful in deriving enhanced analytical specificity and chemical identities from LC-MS data, data set inflation and quantitative imprecision among related features is problematic for statistical analyses and interpretation. This Perspective provides a critical evaluation of global profiling data fidelity with respect to measurement linearity and the quantitative response variation observed among components of the spectra. These elements of data quality are widely overlooked in untargeted metabolomics yet essential for the generation of data that accurately reflect the metabolome. Advanced feature filtering informed by linear range estimation and analyte response factor assessment is advocated as an attainable means of controlling LC-MS data quality in global profiling studies and exemplified herein at both the feature and data set level.

Evaluation of UV-C Decontamination of Clinical Tissue Sections for Spatially Resolved Analysis by Mass Spectrometry Imaging (MSI).

Clinical tissue specimens are often unscreened, and preparation of tissue sections for analysis by mass spectrometry imaging (MSI) can cause aerosolization of particles potentially carrying an infectious load. We here present a decontamination approach based on ultraviolet-C (UV-C) light to inactivate clinically relevant pathogens such as herpesviridae, papovaviridae human immunodeficiency virus, or SARS-CoV-2, which may be present in human tissue samples while preserving the biodistributions of analytes within the tissue. High doses of UV-C required for high-level disinfection were found to cause oxidation and photodegradation of endogenous species. Lower UV-C doses maintaining inactivation of clinically relevant pathogens to a level of increased operator safety were found to be less destructive to the tissue metabolome and xenobiotics. These doses caused less alterations of the tissue metabolome and allowed elucidation of the biodistribution of the endogenous metabolites. Additionally, we were able to determine the spatially integrated abundances of the ATR inhibitor ceralasertib from decontaminated human biopsies using desorption electrospray ionization-MSI (DESI-MSI).

Deep Learning-Based Annotation Transfer between Molecular Imaging Modalities: An Automated Workflow for Multimodal Data Integration.

An ever-increasing array of imaging technologies are being used in the study of complex biological samples, each of which provides complementary, occasionally overlapping information at different length scales and spatial resolutions. It is important to understand the information provided by one technique in the context of the other to achieve a more holistic overview of such complex samples. One way to achieve this is to use annotations from one modality to investigate additional modalities. For microscopy-based techniques, these annotations could be manually generated using digital pathology software or automatically generated by machine learning (including deep learning) methods. Here, we present a generic method for using annotations from one microscopy modality to extract information from complementary modalities. We also present a fast, general, multimodal registration workflow [evaluated on multiple mass spectrometry imaging (MSI) modalities, matrix-assisted laser desorption/ionization, desorption electrospray ionization, and rapid evaporative ionization mass spectrometry] for automatic alignment of complex data sets, demonstrating an order of magnitude speed-up compared to previously published work. To demonstrate the power of the annotation transfer and multimodal registration workflows, we combine MSI, histological staining (such as hematoxylin and eosin), and deep learning (automatic annotation of histology images) to investigate a pancreatic cancer mouse model. Neoplastic pancreatic tissue regions, which were histologically indistinguishable from one another, were observed to be metabolically different. We demonstrate the use of the proposed methods to better understand tumor heterogeneity and the tumor microenvironment by transferring machine learning results freely between the two modalities.

Implications of Peak Selection in the Interpretation of Unsupervised Mass Spectrometry Imaging Data Analyses.

Mass spectrometry imaging can produce large amounts of complex spectral and spatial data. Such data sets are often analyzed with unsupervised machine learning approaches, which aim at reducing their complexity and facilitating their interpretation. However, choices made during data processing can impact the overall interpretation of these analyses. This work investigates the impact of the choices made at the peak selection step, which often occurs early in the data processing pipeline. The discussion is done in terms of visualization and interpretation of the results of two commonly used unsupervised approaches: t-distributed stochastic neighbor embedding and k-means clustering, which differ in nature and complexity. Criteria considered for peak selection include those based on hypotheses (exemplified herein in the analysis of metabolic alterations in genetically engineered mouse models of human colorectal cancer), particular molecular classes, and ion intensity. The results suggest that the choices made at the peak selection step have a significant impact in the visual interpretation of the results of either dimensionality reduction or clustering techniques and consequently in any downstream analysis that relies on these. Of particular significance, the results of this work show that while using the most abundant ions can result in interesting structure-related segmentation patterns that correlate well with histological features, using a smaller number of ions specifically selected based on prior knowledge about the biochemistry of the tissues under investigation can result in an easier-to-interpret, potentially more valuable, hypothesis-confirming result. Findings presented will help researchers understand and better utilize unsupervised machine learning approaches to mine high-dimensionality data.