Fig. 1

Study design for detection of OC. A diagnostic model based on low-pass WGS was used to identify ctDNA from plasma cfDNA using machine learning methods. 159 participants were randomly split into a training cohort (N = 90) and test cohort (N = 69).OC,ovarian cancer; HC, healthy control; WGS, whole genome sequencing; NF, Nucleosome Footprint; Lasso, Least absolute shrinkage and selection operator; SVM, Support Vector Machine