Molecular origin of blood-based infrared fingerprints.

Previous studies demonstrated that infrared absorption spectra of blood sera may help disease detection. For clinical translation of this approach, it is important to determine the molecular origin of disease-related spectral perturbations. To that end, we supplemented infrared spectroscopy with biochemical fractionation and proteomic profiling, which provide detailed information about blood serum composition. We built a model to describe serum absorption based on the concentrations of the highly-abundant proteins and applied this framework to lung cancer detection. We find that it is the levels of acute-phase proteins that change most in the presence of the disease and generate its infrared signature. These findings inform future clinical trials and establish a framework that could be applied to probing of any disease.