Multi-modal breath measurements for biomarker discovery.

Breath contains numerous classes of compounds and biomolecules that could potentially be used as biomarkers for infectious disease as well as a range of other respiratory conditions or states. The goal of this work was to develop a testbed for simultaneous, multi-modal breath measurements. To validate the capabilities of this testbed, a pilot human-subjects research study was conducted to gather a wide range of correlated breath measurements. Seventeen healthy subjects provided breath samples at baseline respiratory rate for particle size, lipid composition and bacterial nucleic acid composition analysis. The majority of the particles the participants exhaled at baseline were smaller than 5μm, consistent with previous literature. A deviation from baseline was detected in one participant immediately prior to COVID-19 symptom onset. This feature persisted for weeks after infection. The exhaled breath particulate contained lipids found in lung surfactant, indicating origin in the lung. Although bacterial DNA was not significantly higher in the exhaled breath particulate than in the environmental background, the metagenome of the breath was distinct from the environment, oral cavity and nasal passages of the participants. The low abundance of the breath microbiome limited analysis. No assertions of statistical significance are offered due to the limited nature of the study scope, the multi-modal breath testbed has promise for discovery of breath biomarkers and as a reference for biomarkers of different classes that are currently being used.