Explainable AI to Facilitate Understanding of Neural Network-Based Metabolite Profiling Using NMR Spectroscopy

Hayden Johnson,Aaryani Tipirneni-Sajja

DOI: https://doi.org/10.3390/metabo14060332

IF: 4.1

2024-06-15

Metabolites

Abstract:Neural networks (NNs) are emerging as a rapid and scalable method for quantifying metabolites directly from nuclear magnetic resonance (NMR) spectra, but the nonlinear nature of NNs precludes understanding of how a model makes predictions. This study implements an explainable artificial intelligence algorithm called integrated gradients (IG) to elucidate which regions of input spectra are the most important for the quantification of specific analytes. The approach is first validated in simulated mixture spectra of eight aqueous metabolites and then investigated in experimentally acquired lipid spectra of a reference standard mixture and a murine hepatic extract. The IG method revealed that, like a human spectroscopist, NNs recognize and quantify analytes based on an analyte's respective resonance line-shapes, amplitudes, and frequencies. NNs can compensate for peak overlap and prioritize specific resonances most important for concentration determination. Further, we show how modifying a NN training dataset can affect how a model makes decisions, and we provide examples of how this approach can be used to de-bug issues with model performance. Overall, results show that the IG technique facilitates a visual and quantitative understanding of how model inputs relate to model outputs, potentially making NNs a more attractive option for targeted and automated NMR-based metabolomics.

biochemistry & molecular biology

What problem does this paper attempt to address?

This paper mainly discusses how to use Explainable Artificial Intelligence (XAI) to enhance understanding of the analysis of nuclear magnetic resonance (NMR) spectroscopy metabolites based on neural networks. In the study, the authors implemented an interpretable AI algorithm called Integrated Gradients (IG) to determine which regions of the input spectrum are most important for the quantification of specific compounds. First, the method was validated in simulated mixture spectra, and then it was applied to the spectra of lipid standard mixtures and mouse liver extracts obtained in experiments. The paper points out that although neural networks have the potential for rapid and large-scale quantification of metabolites in NMR spectroscopy, their non-linear nature makes it difficult to understand the model's prediction behavior. Through the IG method, the study found that the neural network, like a human spectroscopist, identifies and quantifies compounds based on resonance line shapes, amplitudes, and frequencies. Additionally, the neural network can compensate for overlapping peaks and prioritize resonances that are most important for concentration determination. The paper also demonstrates how modifying the neural network training dataset can influence model decisions and provides examples of debugging model performance issues. The research results show that IG technology can promote visual and quantitative understanding of the relationship between model inputs and outputs, making neural networks a more attractive choice for targeted and automated NMR metabolomics.

Explainable AI to Facilitate Understanding of Neural Network-Based Metabolite Profiling Using NMR Spectroscopy

MRSNet: Metabolite Quantification from Edited Magnetic Resonance Spectra With Convolutional Neural Networks

Neural Networks for Conversion of Simulated NMR Spectra from Low-Field to High-Field for Quantitative Metabolomics

An Efficient Approach for Obtaining Small and Macro-molecular 1H NMR Spectra Based on Neural Network

The General Explanation Method with NMR Spectroscopy Enables the Identification of Metabolite Profiles Specific for Normal and Tumor Cell Lines

Using Graph Neural Networks for Mass Spectrometry Prediction

Using neural networks to obtain NMR spectra of both small and macromolecules from blood samples in a single experiment

XAI-enabled neural network analysis of metabolite spatial distributions

Quantitative evaluation of explainable graph neural networks for molecular property prediction

A Bayesian Model of NMR Spectra for the Deconvolution and Quantification of Metabolites in Complex Biological Mixtures

The development of machine learning approaches in two-dimensional NMR data interpretation for metabolomics applications

Unveiling Molecular Moieties through Hierarchical Graph Explainability

XInsight: Revealing Model Insights for GNNs with Flow-based Explanations

Simultaneous Quantitative Analysis of Multiple Metabolites Using Label-Free Surface-Enhanced Raman Spectroscopy and Explainable Deep Learning

Magnetic Resonance Spectroscopy Quantification Aided by Deep Estimations of Imperfection Factors and Macromolecular Signal

Deep Learning-Based Method for Compound Identification in NMR Spectra of Mixtures

An Ensemble Spectral Prediction (ESP) model for metabolite annotation

The application of artificial neural networks in metabolomics: a historical perspective

Automated Machine Learning and Explainable AI (AutoML-XAI) for Metabolomics: Improving Cancer Diagnostics

Benchtop volatilomics supercharged: How machine learning based design of experiment helps optimizing untargeted GC-IMS gas phase metabolomics

Explainable AI: A review of applications to neuroimaging data