Skip to main content
Fig. 2 | BMC Bioinformatics

Fig. 2

From: CancerNet: a unified deep learning network for pan-cancer diagnostics

Fig. 2

Misclassification rates for 4 cancer types selected to illustrate trends observed in CancerNet. A COAD misclassifies primarily to READ with fewer misclassifications in ESCA and STAD. B ESCA misclassifies to HNSC, LUSC and STAD. Lung misclassifications occur often among some sample types. C OV samples misclassify as the two uterine cancer types considered in CancerNet: UCEC and UCS. D LIHC misclassifies as CHOL, MESO, SKCM and NORM. Refer to Abbreviations for cancer types indicated

Back to article page