Skip to main content

Table 5 Average NMI scores on 10 experiments compared with other methods on mus musculus dataset

From: Combining gene ontology with deep neural networks to enhance the clustering of single cell RNA-Seq data

Number of clusters

2

4

6

8

avg1

10

12

avg2

NN(dense)

0.9348

0.8794

0.8179

0.8171

0.8623

/

/

/

NN(ppi/tf)

0.9083

0.8673

0.8119

0.811

0.8496

/

/

/

GONN

0.9688

0.9366

0.8721

0.8635

0.9103

/

/

/

PCA2

0.9527

0.727

0.673

0.6756

0.7571

0.6567

0.6408

0.7210

ICA2

0.8374

0.6966

0.6635

0.6828

0.7201

0.6632

0.6553

0.7261

t-SNE2

0.4025

0.6101

0.6778

0.734

0.6061

0.7432

0.7531

0.6467

PCA10

0.9527

0.7574

0.7349

0.758

0.8008

0.7425

0.7382

0.7835

ICA10

0.1367

0.6224

0.6095

0.7196

0.5221

0.709

0.6965

0.5737

PCA100

0.8656

0.8509

0.7675

0.7812

0.8163

0.7794

0.8215

0.8118

ICA100

0.2186

0.1319

0.1173

0.142

0.15245

0.1539

0.249

0.1665

ZIFA

0.0548

0.3721

0.6267

0.7716

0.4563

0.8188

0.8271

0.5611

pcaReduce

0.6645

0.7115

0.6853

0.6649

0.6816

0.6499

0.6247

0.6689

DAE

0.9621

0.8877

0.8172

0.8194

0.8716

0.8047

0.7957

0.8478

GOAE

0.9544

0.9076

0.8693

0.8545

0.8965

0.8206

0.8018

0.8680

  1. The highest values are shown in blodface. See Table 4 for the more details