Fig. 3: Overview of Fragmentomic Metric Performance in the GRAIL Cohort.
From: Analysis of cfDNA fragmentomics metrics and commercial targeted sequencing panels
Various fragmentomics metrics from WGS studies were tested using a published targeted exon panel for their ability to predict three different cancer types along with healthy vs. cancer using a machine learning model with 10-fold cross validation. Twenty-five replicates were performed and boxplots of the AUROC for each metric are shown. Subtype information was not available for this cohort. Boxplots display the center as the median, with the bounds of the box as Q1 (25th percentile) and Q3 (75th percentile). Whiskers are defined by the lowest and highest value within 1.5 times the interquartile range (IQR; Q3 - Q1). Points outside of 1.5xIQR are displayed as individual points outside of the boxplot.
