Fig. 3
From: Associating somatic mutations to clinical outcomes: a pan-cancer study of survival time
ITH simulation results when the true model contains E. The x-axis denotes the mean sequencing depth. The y-axis denotes the bias of parameter estimates of regression coefficients (βE or βH) and power at α=0.05. Dotted lines denote the bias/power when ITH is known and serve as a benchmark against the estimated ITH metric. H is estimated by PhyloWGS (PhyloWGS(H)), PyClone (PyClone(H)), and SMASH (SMASH(H)). E is estimated by PhyloWGS’s optimal tree (PhyloWGS(oE)), SMASH’s optimal entropy (SMASH(oE)), and SMASH’s weighted entropy (SMASH(wE))