Skip to main content

Table 2 Performance of model training without gene weight layer in different r-radius

From: SWnet: a deep learning model for drug response prediction from cancer genomic signatures and compound chemical structures

Features

r-radius

MSE

\(R^2\)

Gene expression

1

1.0765

0.8489

2

0.9727

0.8634

3

1.0763

0.8489

Gene expression + genetic mutation

1

1.0663

0.8503

2

0.9853

0.8616

3

1.1061

0.8447

  1. Bold values represents the result is the best performance among the models participating in the comparison