Skip to main content
. 2022 Apr 29;16(14):2693–2709. doi: 10.1002/1878-0261.13210

Fig. 5.

Fig. 5

Clinical implications. (A) Cox proportional hazards model for the effect on overall survival (OS) for the 10 most important miRaCl/miRaCl‐20 features on EGAS1127 dataset, error bars represent 95% confidence intervals. (B, C) Kaplan–Meier analyses for OS stratified by the miRaCl or miRaCl‐20 predicted CMS classes in EGAS1127 dataset (n = 82), log‐rank P‐value from score test. (D) Comparison of miRaCl‐predicted CMS classes in patients with paired primary (prim.) and metastasis (met.) samples, including one recurrent (recur.) colorectal tumor; LN, lymph node; NOS, not otherwise specified. If multiple metastases were available, the primary was duplicated for visualization purposes, as marked by underscore extension of patient IDs (n = 38).