BMC Bioinformatics

Table 4 Common features identified across runs for the different omics data types

From: Novel deep learning-based solution for identification of prognostic subgroups in liver cancer (Hepatocellular carcinoma)

Omics type	No. of runs	Loss functions
Omics type	No. of runs	BCE	MSE	L_RC	L_RS	L_RSC
mRNA	6 runs	1296	1319	1268	422	888
	8 runs	929	986	742	256	633
	10 runs	451	295	439	124	377
miRNA	6 runs	31	34	36	9	16
	8 runs	22	21	20	1	9
	10 runs	12	1	15	1	3
Methylation	6 runs	1676	1593	1794	647	1135
	8 runs	1252	1203	1388	330	856
	10 runs	675	520	1097	89	434

A summary of the common features identified for the different omics data types (mRNA, miRNA, methylation) across six, eight and ten replicate runs of models with the five different losses

Back to article page

ISSN: 1471-2105

Contact us

General enquiries: journalsubmissions@springernature.com