Skip to main content

Table 3 Mean AUROC on external test set

From: A fair experimental comparison of neural network architectures for latent representations of multi-omics for drug response prediction

Drug

Omics stacking

MOLI

Super.FELT

Early integration

OmiEmbed

MOMA

PCA

Gemcitabine TCGA

0.655 ± 0.029

0.640 ± 0.037

0.618 ± 0.042

0.604 ± 0.093

0.565 ± 0.059

0.473 ± 0.039

0.527 ± 0.085

Gemcitabine PDX

0.714 ± 0.089

0.614 ± 0.044

0.692 ± 0.054

0.525 ± 0.099

0.657 ± 0.119

0.627 ± 0.092

0.600 ± 0.088

Cisplatin

0.644 ± 0.087

0.674 ± 0.032

0.728 ± 0.045

0.604 ± 0.052

0.513 ± 0.056

0.687 ± 0.019

0.743 ± 0.056

Docetaxel

0.584 ± 0.101

0.647 ± 0.038

0.588 ± 0.056

0.456 ± 0.065

0.478 ± 0.056

0.581 ± 0.064

0.475 ± 0.122

Erlotinib

0.744 ± 0.065

0.722 ± 0.127

0.563 ± 0.080

0.789 ± 0.079

0.633 ± 0.105

0.715 ± 0.132

0.800 ± 0.078

Cetuximab

0.575 ± 0.049

0.476 ± 0.111

0.556 ± 0.099

0.470 ± 0.130

0.468 ± 0.075

0.505 ± 0.019

0.523 ± 0.134

Paclitaxel

0.619 ± 0.152

0.547 ± 0.121

0.527 ± 0.114

0.418 ± 0.062

0.516 ± 0.026

0.573 ± 0.124

0.435 ± 0.104

Mean Rank

2.14

3.43

3.14

5.57

5.43

4.14

4.14

  1. Best results are shown in bold and second best are italic. The values represent the means and standard deviations over five iterations