NMRVPF: A High-Precision Method for Automatic Identification and Quantitative Analysis of Mixtures Based on 1D NMR Spectra

Nuclear magnetic resonance (NMR) is a powerful analytical tool for identifying and quantifying components in complex mixtures. However, analyzing crowded spectra is challenging due to signal overlap and variations in chemical shifts, intensities, line widths, and line shapes across different samples. These issues are further compounded in metabolite-rich systems, where severe spectral crowding limits the quantification accuracy. In this study, we present a robust approach using 1D 1 H NMR spectroscopy to address these challenges, with amino acid mixtures serving as a model system. Our method constructs a comprehensive peak feature library that incorporates multidimensional descriptors (chemical shifts, intensities, line widths) and employs iterative optimization techniques for precise spectral fitting. By dynamically adjusting spectral parameters, the reconstructed spectrum achieves near-perfect matching (MSE < 0.05) with the experimental spectrum, even in densely crowded regions. This work demonstrates the potential of 1D NMR spectroscopy as a versatile and reliable tool for analyzing complex mixtures, overcoming key challenges in metabolomics, pharmaceutical quality control, and chemical biology studies.