Table 1.
Feature selection methods implemented in different scRNA-seq clustering algorithms
| scRNA-seq clustering method | Reference | Quantities used for feature selection |
|---|---|---|
| Seurat | [10] |
and 
|
| PanoView | [23] |
and  (similar to Seurat) |
| SC3 | [9] |
and 
|
| Monocle | [24] |
and 
|
| SCANPY | [25] |
|
| scVI | [26] |
|
| TSCAN | [11] |
and 
|
| SAIC | [27] | loess regression between  and 
|
| SCENT | [28] |
|
| SOUP | [29] | Gini index and 
|
| FiRE | [32] |
and 
|
| SINCERA | [33] |
,  and cell specificity index |
| RaceID3 | [34] | Second-order polynomial between  and 
|
Mean is denoted as 
Variance is denoted as 
Dispersion is denoted as 
Coefficient of variation is denoted as 
Dropout rate is denoted as 
SPCA means the sparse PCA algorithm. SVD means the singular value decomposition.