Strategies for MCR image analysis of large hyperspectral data-sets

Applications of multivariate analysis and unsupervised machine learning to ToF-SIMS images of organic, bioorganic, and biological systems

Time-of-flight secondary ion mass spectrometry (ToF-SIMS) imaging offers a powerful, label-free method for exploring organic, bioorganic, and biological systems. The technique is capable of very high spatial resolution, while also producing an enormous amount of information about the chemical and molecular composition of a surface. However, this information is inherently complex, making interpretation and analysis of the vast amount of data produced by a single ToF-SIMS experiment a considerable challenge. Much research over the past few decades has focused on the application and development of multivariate analysis (MVA) and machine learning (ML) techniques that find meaningful patterns and relationships in these datasets. Here, we review the unsupervised algorithms—that is, algorithms that do not require ground truth labels—that have been applied to ToF-SIMS images, as well as other algorithms and approaches that have been used in the broader family of mass spectrometry imaging (MSI) techniques. We first give a nontechnical overview of several commonly used classes of unsupervised algorithms, such as matrix factorization, clustering, and nonlinear dimensionality reduction. We then review the application of unsupervised algorithms to various organic, bioorganic, and biological systems including cells and tissues, organic films, residues and coatings, and spatially structured systems such as polymer microarrays. We then cover several novel algorithms employed for other MSI techniques that have received little attention from ToF-SIMS imaging researchers. We conclude with a brief outline of potential future directions for the application of MVA and ML algorithms to ToF-SIMS images.

High-Throughput and Combinatorial Approaches for the Development of Multifunctional Polymers

High-throughput (HT) development of new multifunctional polymers is accomplished by the combination of different HT tools established in polymer sciences in the last decade. Important advances are robotic/HT synthesis of polymer libraries, the HT characterization of polymers, and the application of spatially resolved polymer library formats, explicitly microarray and gradient libraries. HT polymer synthesis enables the generation of material libraries with combinatorial design motifs. Polymer composition, molecular weight, macromolecular architecture, etc. may be varied in a systematic, fine-graded manner to obtain libraries with high chemical diversity and sufficient compositional resolution as model systems for the screening of these materials for the functions aimed. HT characterization allows a fast assessment of complementary properties, which are employed to decipher quantitative structure-properties relationships. Moreover, these methods facilitate the HT determination of important surface parameters by spatially resolved characterization methods, including time-of-flight secondary ion mass spectrometry and X-ray photoelectron spectroscopy. Here current methods for the high-throughput robotic synthesis of multifunctional polymers as well as their characterization are presented and advantages as well as present limitations are discussed.

Understanding mass spectrometry images: complexity to clarity with machine learning

The application of artificial intelligence and machine learning to hyperspectral mass spectrometry imaging (MSI) data has received considerable attention over recent years. Various methodologies have shown great promise in their ability to handle the complexity and size of MSI data sets. Advances in this area have been particularly appealing for MSI of biological samples, which typically produce highly complicated data with often subtle relationships between features. There are many different machine learning approaches that have been applied to MSI data over the past two decades. In this review, we focus on a subset of non-linear machine learning techniques that have mostly only been applied in the past 5 years. Specifically, we review the use of the self-organizing map (SOM), SOM with relational perspective mapping (SOM-RPM), t-distributed stochastic neighbor embedding (t-SNE) and uniform manifold approximation and projection (UMAP). While not their only functionality, we have grouped these techniques based on their ability to produce what we refer to as similarity maps. Similarity maps are color representations of hyperspectral data, in which spectral similarity between pixels-that is, their distance in high-dimensional space-is represented by relative color similarity. In discussing these techniques, we describe, briefly, their associated algorithms and functionalities, and also outline applications in MSI research with a strong focus on biological sample types. The aim of this review is therefore to introduce this relatively recent paradigm for visualizing and exploring hyperspectral MSI, while also providing a comparison between each technique discussed.

Discriminating normal regions within cancerous hen ovarian tissue using multivariate hyperspectral image analysis

RATIONALE

Identification of subregions under different pathological conditions on cancerous tissue is of great significance for understanding cancer progression and metastasis. Infrared matrix-assisted laser desorption electrospray ionization mass spectrometry (IR-MALDESI-MS) can be potentially used for diagnostic purposes since it can monitor spatial distribution and abundance of metabolites and lipids in biological tissues. However, the large size and high dimensionality of hyperspectral data make analysis and interpretation challenging. To overcome these barriers, multivariate methods were applied to IR-MALDESI data for the first time, aiming at efficiently resolving mass spectral images, from which these results were then used to identify normal regions within cancerous tissue.

METHODS

Molecular profiles of healthy and cancerous hen ovary tissues were generated by IR-MALDESI-MS. Principal component analysis (PCA) combined with color-coding built a single tissue image which summarizes the high-dimensional data features. Pixels with similar color indicated similar composition. PCA results from healthy tissue were further used to test each pixel in cancerous tissue to determine if it is healthy. Multivariate curve resolution-alternating least squares (MCR-ALS) was used to obtain major spatial features existing in ovary tissues, and group molecules with the same distribution patterns simultaneously.

RESULTS

PCA as the predominating dimensionality reduction approach captured over 90% spectral variances by the first three PCs. The PCA images show the cancerous tissue is more chemically heterogeneous than healthy tissue, where at least four regions with different m/z profiles can be differentiated. PCA modeling assigns top regions of cancerous tissue as healthy-like. MCR-ALS extracted three and four major compounds from healthy and cancerous tissue, respectively. Evaluating similarities of resolved spectra uncovered the chemical components that were distinct in some regions on cancerous tissue, serving as a supplementary way to differentiate healthy and cancerous regions.

CONCLUSIONS

Two unsupervised chemometric methods including PCA and MCR-ALS were applied for resolving and visualizing IR-MALDESI-MS data acquired from hen ovary tissues, improving the interpretation of mass spectrometry imaging results. Then possible normal regions were differentiated from cancerous tissue sections. No prior knowledge is required using either chemometric method, so our approach is readily suitable for unstained tissue samples, which allows one to reveal the molecular events happening during disease progression.

A Time-of-Flight Secondary Ion Mass Spectrometry/Multivariate Analysis (ToF-SIMS/MVA) Approach To Identify Phase Segregation in Blends of Incompatible but Extremely Similar Resins

This work presents a data analysis extension to a well-established methodology for the assessment of organic coatings using imaging time-of-flight secondary ion mass spectrometry (ToF-SIMS). Such an approach produced results that can be analyzed using a multivariate analysis (MVA) procedure that performs the simultaneous processing of spatially and chemically related datasets. The coatings consist of two commercial resins that yield extremely similar spectra, and there are no peaks of sufficient intensity that are uniquely diagnostic of either material to provide an unambiguous identification of each. In order to resolve the problem, in addition to microtome-based sample preparation steps of tapers for the analysis through sample thickness, standard samples in cured and uncured conditions are introduced and measured in the same fashion as the specimens under investigation. The resulting ToF-SIMS imaging datasets have been processed using non-negative matrix factorization (NMF), which enabled identification of phase separation in the cured coatings.

The physicochemical fingerprint of Necator americanus

Necator americanus, a haematophagous hookworm parasite, infects ~10% of the world's population and is considered to be a significant public health risk. Its lifecycle has distinct stages, permitting its successful transit from the skin via the lungs (L3) to the intestinal tract (L4 maturing to adult). It has been hypothesised that the L3 larval sheath, which is shed during percutaneous infection (exsheathment), diverts the immune system to allow successful infection and reinfection in endemic areas. However, the physicochemical properties of the L3 larval cuticle and sheath, which are in direct contact with the skin and its immune defences, are unknown. In the present study, we controlled exsheathment, to characterise the sheath and underlying cuticle surfaces in situ, using atomic force microscopy (AFM) and time-of-flight secondary ion mass spectrometry (ToF-SIMS). AFM revealed previously unseen surface area enhancing nano-annuli exclusive to the sheath surface and confirmed greater adhesion forces exist between cationic surfaces and the sheath, when compared to the emergent L3 cuticle. Furthermore, ToF-SIMS elucidated different chemistries between the surfaces of the cuticle and sheath which could be of biological significance. For example, the phosphatidylglycerol rich cuticle surface may support the onward migration of a lubricated infective stage, while the anionic and potentially immunologically active heparan sulphate rich deposited sheath could result in the diversion of immune defences to an inanimate antigenic nidus. We propose that our initial studies into the surface analysis of this hookworm provides a timely insight into the physicochemical properties of a globally important human pathogen at its infective stage and anticipate that the development and application of this analytical methodology will support translation of these findings into a biological context.

ToF-SIMS analysis of a polymer microarray composed of poly(meth)acrylates with C6 derivative pendant groups

Surface analysis plays a key role in understanding the function of materials, particularly in biological environments. Time‐of‐flight secondary ion mass spectrometry (ToF‐SIMS) provides highly surface sensitive chemical information that can readily be acquired over large areas and has, thus, become an important surface analysis tool. However, the information‐rich nature of ToF‐SIMS complicates the interpretation and comparison of spectra, particularly in cases where multicomponent samples are being assessed. In this study, a method is presented to assess the chemical variance across 16 poly(meth)acrylates. Materials are selected to contain C₆ pendant groups, and ten replicates of each are printed as a polymer microarray. SIMS spectra are acquired for each material with the most intense and unique ions assessed for each material to identify the predominant and distinctive fragmentation pathways within the materials studied. Differentiating acrylate/methacrylate pairs is readily achieved using secondary ions derived from both the polymer backbone and pendant groups. Principal component analysis (PCA) is performed on the SIMS spectra of the 16 polymers, whereby the resulting principal components are able to distinguish phenyl from benzyl groups, mono‐functional from multi‐functional monomers and acrylates from methacrylates. The principal components are applied to copolymer series to assess the predictive capabilities of the PCA. Beyond being able to predict the copolymer ratio, in some cases, the SIMS analysis is able to provide insight into the molecular sequence of a copolymer. The insight gained in this study will be beneficial for developing structure–function relationships based upon ToF‐SIMS data of polymer libraries. © 2016 The Authors Surface and Interface Analysis Published by John Wiley & Sons Ltd.

Analysis and prediction of defects in UV photo-initiated polymer microarrays

Polymer microarrays are a key enabling technology for the discovery of novel materials. This technology can be further enhanced by expanding the combinatorial space represented on an array. However, not all materials are compatible with the microarray format and materials must be screened to assess their suitability with the microarray manufacturing methodology prior to their inclusion in a materials discovery investigation. In this study a library of materials expressed on the microarray format are assessed by light microscopy, atomic force microscopy and time-of-flight secondary ion mass spectrometry to identify compositions with defects that cause a polymer spot to exhibit surface properties significantly different from a smooth, round, chemically homogeneous 'normal' spot. It was demonstrated that the presence of these defects could be predicted in 85% of cases using a partial least square regression model based upon molecular descriptors of the monomer components of the polymeric materials. This may allow for potentially defective materials to be identified prior to their formation. Analysis of the PLS regression model highlighted some chemical properties that influenced the formation of defects, and in particular suggested that mixing a methacrylate and an acrylate monomer and/or mixing monomers with long and linear or short and bulky pendant groups will prevent the formation of defects. These results are of interest for the formation of polymer microarrays and may also inform the formulation of printed polymer materials generally.