Table 1 Summary of the features used in the classification task
From: An ensemble learning approach for modeling the systems biology of drug-induced injury
Type of feature | Name | Description |
|---|---|---|
Gene expression features | Landmark genes | 978 genes directly measured from the L1000 datasets |
DisGeNET DILI genes | Curated genes associated to 9 phenotypes related with DILI from DisGeNET database | |
GUILDify DILI genes | Genes associated through the protein interactions network to 6 phenotypes related with DILI using GUILDify | |
DILI landmark genes | 66 landmark genes selected using non-parametric test for each gene across all samples of Most/Less- vs. No-DILI-Concern drugs (P-value< 0.05) | |
Structural features | SMILES | Line notation describing the chemical structure of drugs |
Drug target genes | Set of targets | 1664 drug targets retrieved from DGIdb, HitPick and SEA |