Automatic Identification of Potential Cellular Metabolites for Untargeted NMR Metabolomics

Area of Science:

• Metabolomics
• Biochemistry
• Bioinformatics

Background:

• Metabolic profiles offer insights into organismal physiology and pathology.
• Nuclear Magnetic Resonance (NMR) spectroscopy non-invasively monitors metabolite changes in biological samples.
• Proton NMR spectral overlap in complex samples hinders metabolite identification and monitoring.

Purpose of the Study:

• To introduce ROIAL-NMR, a Python program for systematic metabolite identification from proton NMR spectral regions-of-interest (ROIs).
• To utilize the Human Metabolome Database (HMDB) as a reference for metabolite identification in complex biological samples.
• To differentiate disease states (e.g., lung cancer, Alzheimer's disease and related dementia) from controls using spectral intensity variations.

Main Methods:

• Development of a Python program, ROIAL-NMR, to analyze proton NMR spectra.
• Defining spectral regions-of-interest (ROIs) based on sample types and disease conditions.
• Utilizing the Human Metabolome Database (HMDB) for metabolite identification and comparison.

Main Results:

• The ROIAL-NMR program successfully identified 88 potential metabolites from complex biological samples.
• The program distinguished 66 metabolites differentiating lung cancer (LC) patients from controls.
• It identified 80 metabolites differentiating LC patients with Alzheimer's disease and related dementia (ADRD) from those without ADRD.

Conclusions:

• ROIAL-NMR provides a systematic approach to identify metabolites in complex biological samples using NMR spectroscopy.
• The program effectively aids in differentiating disease states based on metabolic profiles.
• This tool offers valuable insights into the pathophysiology of diseases like lung cancer and ADRD.