Assessment of deep learning and transfer learning for cancer prediction based on gene expression data

BMC Bioinformatics

Table 1 Characteristics of the microarray dataset

Disease	Size	Patients	Cell lines	Cancer	Non-cancer	Prior
Leukemias	4283	3452	831	2336	1947	0.55
Bone marrow cancer	3525	3374	151	3185	340	0.90
Breast cancer	2171	1366	805	1863	308	0.86
Kidney cancer	657	423	234	400	257	0.61
Liver cancer	727	312	415	601	126	0.82
Lung cancer	1415	749	666	818	597	0.58
Skin cancer	835	554	281	454	381	0.54
Brain cancer	869	468	401	819	50	0.94
Colon cancer	1239	875	364	1112	127	0.90
Ovary cancer	573	427	146	533	40	0.93
Prostate cancer	415	182	233	350	65	0.84
Total	16,709	12,182	4527	12,471	4238	0.75

The columns represent respectively the type of tissues (Disease), the numbers of samples (Size), patient samples (Patients), cell line samples (Cell lines), cancer samples (Cancer), non-cancer samples (Non-cancer) and the proportion of the majority class (Prior)

ISSN: 1471-2105