The compositional space of exhaled breath condensate and its link to the human breath volatilome

Breath analysis is commonly understood to target gaseous or volatile organic compounds (VOCs) for the characterization of different pathologies. Targeted analysis is most effective if a working hypothesis can be based on a plethora of data. The recently published volatilome builds an optimal basis for organizing powerful target sets. However, the origin and pathways of biosynthesis of many VOCs are not known, which complicates the formulation of useful hypotheses. To find the missing link between VOCs and their origin, it is necessary to analyze their precursor fluids themselves. In order to provide condensation nuclei for the generation of future hypotheses, we provide the compositional space over 23 samples of the unperturbed human exhaled breath condensate (EBC) metabolome. We propose a way to connect the compositional spaces of both VOCs and EBC so as to gain insight into the most probable form of VOC precursors. In a way analogous to tandem MS it is possible to create a mass difference network over compositional data by linking compositions with mass differences that are designed to mimic biochemical reactions. We propose to use mass difference enrichment analysis (MDEA) in order to mine probable relations between VOCs and their precursor fluids. We have found 2691 EBC compositions and linked them to 235 breath VOC compositions that correspond to 848 individual compounds. We found that VOCs are likely to be found as hexose conjugates or as amino acid conjugates with Glutamine or Asparagine playing a major role. Furthermore, we found that dicarboxylic acid mass differences may be more indicative for oxidative stress than oxygenation-hydrogenation sequences.