The B-PC2 network has four genes associated with the left branch leading to Leaf 1. While S100A10 and EIF4A1 are involved in more general cellular functions, IGHM and TCL1A are specific to T cell maturation and adaptive immune response. Equally compelling is the association of PRELID1 with the right branch of B-PC2 where inhibition of Th2 cell development potentially reduces the contribution of adaptive LY2157299 immunity to asthma ; subjects in Leaf 1 where Th2 cell-activated eosinophils and lymphocytes dominate have the complementary absence of any PRELID1-induced protection. The remaining genes associated with the right branch of B-PC2 relate to chemokine signaling and immune cell activation. C-PC2 Vorinostat separates the subjects from the right branch of B-PC2 into an asthmatic endotype Leaf 2 characterized by lower allergen specific IgE levels and Leaf 3 characterized by high atopy with low asthma prevalence. The network for the C-PC2 metagene contains three informative hub nodes which potentially play a prominent role in the biological processes underlying the observed gene expression changes. The immunoglobulin hub node suggests underlying adaptive immune responses for the Leaf 2 asthma endotype. A second hub is an NFkB complex, broadly associated with enhanced inflammation associated with adaptive and innate immune response. Our novel, multi-step, systems-based decision tree approach using principal components-summarized gene expression and clinical biomarker correlations differentiated asthmatics from non-asthmatics revealing biological pathways that potentially underlie the varied asthmatic endotypes. The results of this study are for children aged 9 to 13 years and cannot be extrapolated to all ages. In adults, for example, there are higher prevalence rates of other asthma phenotypes and endotypes, such as non-allergic asthma and aspirin-exacerbated asthma. Characteristics of the data-driven derived endotypes from this study are consistent with previously published endotypes based solely on clinical diagnostic criteria, but our data-driven method provides mechanistic understanding that is not possible when using established clinical markers alone. One theme that emerges from this analysis is the interplay between innate and adaptive immune responses. We clearly see a dominant role for adaptive immunity in Leaf 1, innate immunity in Leaves 5 and 8, with a mixed contribution in Leaf 2. Our results also suggest a role for broad systemic inflammation in addition to the localized hyperreactivity in the lung as a major driver for asthma.