Skip to main content

Table 1 Optimization of the size of the target local environment in the PPI network

From: GraphDTI: A robust deep learning predictor of drug-target interactions from multiple heterogeneous data

N

AUC

10

0.978 ±0.003

20

0.985 ±0.002

30

0.989 ±0.002

40

0.991 ±0.002

50

0.993 ±0.001

60

0.985 ±0.002

70

0.985 ±0.002

  1. The Area Under the Curve (AUC) measures the classification performance of the MLP model employing drug and local network embeddings in 5-fold cross-validation against the GraphDTI dataset. Graph embeddings for target nodes are calculated for a number of highly confident neighbors of a target node (N) increasing from 10 to 70