Figure 5.
Network analysis of proteins associated with isolated PE. STRING network analysis incorporating proteins with 10% highest or lowest expression within the isolated PE phenotype relative to the global average, proteins identified as specific to the isolated PE phenotype in LASSO-regularized regression, and proteins highly correlated (Spearman’s ρ > 0.7) with these proteins in the sample. The non-networked proteins shown are sorted by strength of evidence, which is based on their consistency of selection across LASSO-regularized regression models: the strongest evidence for an association with isolated PE is provided by the proteins being selected in both the isolated PE vs isolated DVT as well as the isolated PE vs DVT-associated PE models. Node colors indicate cluster membership, as determined using the unsupervised Markov clustering algorithm. Dashed lines indicate inter-cluster edges, whereas solid lines represent intra-cluster connections. Protein–protein connections (ie, the edges of the graph) reflect experimentally established protein–protein interactions, coexpression patterns, protein homology, and putative connections based on text mining.

Network analysis of proteins associated with isolated PE. STRING network analysis incorporating proteins with 10% highest or lowest expression within the isolated PE phenotype relative to the global average, proteins identified as specific to the isolated PE phenotype in LASSO-regularized regression, and proteins highly correlated (Spearman’s ρ > 0.7) with these proteins in the sample. The non-networked proteins shown are sorted by strength of evidence, which is based on their consistency of selection across LASSO-regularized regression models: the strongest evidence for an association with isolated PE is provided by the proteins being selected in both the isolated PE vs isolated DVT as well as the isolated PE vs DVT-associated PE models. Node colors indicate cluster membership, as determined using the unsupervised Markov clustering algorithm. Dashed lines indicate inter-cluster edges, whereas solid lines represent intra-cluster connections. Protein–protein connections (ie, the edges of the graph) reflect experimentally established protein–protein interactions, coexpression patterns, protein homology, and putative connections based on text mining.

Close Modal

or Create an Account

Close Modal
Close Modal