HIGHLIGHTS
SUMMARY
The authors used L1 regularized multivariable logistic regression models (LASSO)43 to select uncorrelated proteins that best predicted COVID-19 severity in the BQC19 training cohort. The authors first defined a baseline model which included only the four covariates in the logistic regression model: age, sex, sample processing time, and hospital site to predict COVID-19 severity. From the lambda parameter search are shown in Supplementary Fig 4A, B. For the best performing model predicting severe COVID-19, the authors selected a l og10 lambda value of - 1.5 which generated an average training . . .
If you want to have access to all the content you need to log in!
Thanks :)
If you don't have an account, you can create one here.