Skip to main content

Table 3 Optimal numbers of cell types identified using different methods for the five real datasets

From: ZINBMM: a general mixture model for simultaneous clustering and gene selection using single-cell transcriptomic data

Method

True

Proposed

CIDR

SC3

Seurat

scDeepCluster

MNN-Graph

MNN-Kmeans

RZiMM

snbClust

sparseKmeans

Dataset

   Mouse embryonic stem cells

3

3

5

5

6

7

4

3

2

2

5

   Human lung cancer cells

3

3

6

12

14

12

2

2

5

4

2

   Mouse uterus cells

8

5

12

11

10

9

7

2

8

7

2

   Mouse liver cells

8

4

12

12

12

11

7

2

7

8

2

   Mouse mammary gland cells

14

5

8

16

15

17

16

2

4

6

2