Classification and severity progression measure of COVID-19 patients using pairs of multi-omic factors

Early detection and effective treatment of severe COVID-19 patients remain two major challenges during the current pandemic. Analysis molecular changes in blood samples is one promising approaches to this problem. From thousands proteomic, metabolomic, lipidomic, transcriptomic biomarkers selected other research, we identify several pairs that after additional nonlinear spline transformation are highly classifying predicting cases. The performance these evaluated in-sample, a cross-validation exercise, an out-of-sample analysis on independent datasets. We further improve our classifier by identifying complementary using hierarchical clustering. In result, achieve 96–98% AUC validation data. Our findings can help medical experts small groups be used construct cost-effective test for patient screening prediction severity progression.