Skip to main content
. Author manuscript; available in PMC: 2022 Apr 8.
Published in final edited form as: IEEE/ACM Trans Comput Biol Bioinform. 2021 Apr 8;18(2):633–643. doi: 10.1109/TCBB.2019.2921577

Table 1.

Summary of the main characteristics of the 20 cancer gene expression benchmark datasets analyzed in our study. The indicated number of classes is that found in the original study.

Cancer type Dataset Objects Features Classes
Skin Bittner 38 110 2
Liver Chen 179 4 2
Lung Golub-V1 72 93 2
Colon Laiho 37 110 2
Breast West 49 60 2
Leukemia Amstrong- 72 110 3
Bladder Dyrskjot 40 60 3
Leukemia Golub-V2 72 93 3
Blood Alizadeh-V2 62 3 3
Brain Liang 37 20 3
Brain Nutt-V1 50 20 4
Lung Garber 66 2 4
Prostate Lapointe-V2 110 3 4
Endometrium Risinger 42 20 4
Prostate Tomlins-V2 92 20 4
Lung Bhattacharjee 203 7 5
Brain Pomeroy-V2 42 69 5
Prostate Tomlins-V1 104 4 5
Bone marrow Yeoh-V2 248 20 6
Multi-tissue Su 174 2 10