CRISP: a deep learning architecture for GC × GC–TOFMS contour ROI identification, simulation and analysis in imaging metabolomics

Abstract

Two-dimensional gas chromatography–time-of-flight mass spectrometry (GC × GC–TOFMS) provides a large amount of molecular information from biological samples. However, the lack of a comprehensive compound library or customizable bioinformatics tool is currently a challenge in GC × GC–TOFMS data analysis. We present an open-source deep learning (DL) software called contour regions of interest (ROI) identification, simulation and untargeted metabolomics profiler (CRISP). CRISP integrates multiple customizable deep neural network architectures for assisting the semi-automated identification of ROIs, contour synthesis, resolution enhancement and classification of GC × GC–TOFMS-based contour images. The approach includes the novel aggregate feature representative contour (AFRC) construction and stacked ROIs. This generates an unbiased contour image dataset that enhances the contrasting characteristics between different test groups and can be suitable for small sample sizes. The utility of the generative models and the accuracy and efficacy of the platform were demonstrated using a dataset of GC × GC–TOFMS contour images from patients with late-stage diabetic nephropathy and healthy control groups. CRISP successfully constructed AFRC images and identified over five ROIs to create a deepstacked dataset. The high fidelity, 512 × 512-pixels generative model was trained as a generator with a Fréchet inception distance of <47.00. The trained classifier achieved an AUROC of >0.96 and a classification accuracy of >95.00% for datasets with and without column bleed. Overall, CRISP demonstrates good potential as a DL-based approach for the rapid analysis of 4-D GC × GC–TOFMS untargeted metabolite profiles by directly implementing contour images. CRISP is available at https://github.com/vivekmathema/GCxGC-CRISP.

Recent Advances in GC×GC and Chemometrics to Address Emerging Challenges in Nontargeted Analysis

“…Hence, image reduction can also be a valuable preprocessing step for pixel-based analysis by balancing computational time and model accuracy. Furthermore, an open-source deep learning software was also recently introduced to discover features and classify samples based on the GC×GC-TOFMS contour images . This methodology was termed contour regions of interest identification, simulation, and untargeted metabolomics profiler (CRISP).…”

“…Furthermore, an open-source deep learning software was also recently introduced to discover features and classify samples based on the GC×GC-TOFMS contour images. 107 This methodology was termed contour regions of interest identification, simulation, and untargeted metabolomics profiler (CRISP). CRISP was demonstrated to classify patients with late-stage diabetic nephropathy from healthy controls with accuracies greater than 96% at different image resolutions.…”

Small molecule metabolites: discovery of biomarkers and therapeutic targets

“… 259 – 269 The larger data sets require the specialized tools for rapid analysis. 270 – 273 The progressions such as automatic annotation, in-silico fragmentation and databases construction have advanced to solving these problems. 274 – 276 Multivariate statistical techniques are widely applied in mechanistic understanding of metabolic processes, beyond phenotyping and biomarker discovery of various diseases.…”

Theoretical modeling and machine learning-based data processing workflows in comprehensive two-dimensional gas chromatography—A review

“…Nevertheless, to ensure that the reader can grasp the concepts behind these algorithms, we provide a brief fundamental understanding of the functioning of the discussed ML techniques. Moreover, it is worth noting the emergence of deep learning algorithms as a promising tool for large dataset analysis [96][97][98]; however, this topic falls outside the scope of this review.…”