Skip to main content
. 2023 Apr 12;10:1123682. doi: 10.3389/fcvm.2023.1123682

Figure 1.

Figure 1

Analysis workflow applied to the CDCS cohort. The background network reflecting the conditional dependence structure among the four data types was inferred using graphical LASSO. From the network, subnetwork signatures of secondary MACE and HF were obtained using the training dataset (CDCS) using 10-fold cross validation for choosing the optimal regularization parameter.