Metabolite analysis in sepsis through conditional independence maps

Sepsis is the response of the host to an infection that produces lesions in its own organs and tissues. Despite the great advances in modern medicine, including vaccines, antibiotics and intensive care, it is still the primary cause of death due to infection. Sepsis may result in shock, multi-organic failure and death unless there is a rapid identification of the infection and timely administration of treatment. Its mortality rates can reach up to 45.7% for septic shock, its most acute manifestation. In this paper we also present these conditional independence maps in the context of algebraic statistics. The results of this analysis over a small cohort of nine patients at three different times (ICU admission, 48h and ICU discharge) showed that there is a significant interaction between C3- DC / C4-OH (Hydroxybutyrylcarnitine) and C5 (Valerylcarnitin) for the three time snapshots. We also found a significant interaction between C3-DC / C4-OH (Hydroxybutyrylcarnitine) and C5 (Valerylcarnitine) and Isoleucine (Ile) at 48h and ICU discharge.