Skip to main content
. 2019 Jan 30;9:1005. doi: 10.1038/s41598-018-37835-6

Figure 2.

Figure 2

Sensitivity and specificity is robust across three different random training-test partitions of available samples. The average difference between any two pairs of partitionings is less than 4.2% for both sensitivity and specificity across all seventeen cancer types. Error bars represent 95% confidence intervals.