Biomolecular Profile of Tobacco Users by Infrared Spectroscopy and Machine Learning Approaches
Author(s): Marcelo Saito Nogueira, Natalia Caroline Queiroz Costa, Kusai Baroudi, Maria Clara de Moura Santos Coelho Ferreira, Sara Maria Santos Dias da Silva, Leonardo Barbosa Leal, Pedro Arthur Augusto Castro, Felipe Peralta, Denise Maria Zezell, Luis Felipe das Chagas e Silva de Carvalho
Objectives: Our study validates the preliminary steps needed to introduce FT-IR spectroscopy as a point-of-care diagnostic tool, particularly for patients at high risk for cancer.
Materials and methods: FT-IR spectroscopy was used to determine molecular changes and classify saliva samples of control, smoker, and occasional smoker groups.
Results: Correctly classified instances were 72.7% for the control group, 65.5% for occasional smokers and 75% for smokers. Sample differences were observed in the peaks at 1076cm-1, 1403cm-1 symmetric CH3 modes of protein methyl groups and δsCH3 of collagen, 1451cm-1 asymmetrical CH3 bending modes of the protein methyl groups, 1547cm-1 of protein band, amide II, peptide and proteins amide II, and 1646cm-1 amide I, C5 methylated cytosine, C==O bond, C==C stretching uracil and NH2 guanine.
Conclusion: Our research demonstrates the potential of FT-IR spectroscopy in detecting subtle molecular changes in saliva, which can be correlated with smoking habits. This non-invasive technique could be instrumental in the early detection and monitoring of oral and systemic diseases, especially those related to tobacco use. Future research should focus on refining the classification algorithms and expanding the sample size to further improve the diagnostic accuracy and reliability of this technique.