X-DILiver


The liver, which plays a critical role in the metabolism of xenobiotics, is highly susceptible to damage from drugs and their metabolites. Thus, there is a demand for an accurate, in silico model for the prediction of drug candidate hepatotoxicity, which would greatly benefit drug development. We developed a drug-induced liver injury (DILI) prediction model trained on a large-scale hepatotoxicity dataset that uses an ensemble strategy. "X-DILiver" can assist successful drug development by accurately predicting the biological safety associated with the hepatotoxicity of drug candidates.

Submission


Input a SMILES string below. To predict the toxicity of multiple molecules simultaneously, a SMILES file with the extension ’.smi’ can be queried.

For this query file, please input SMILES strings line by line.

Input SMILES:

  • Example

  • /

  • /


Submit .SMI File Upload:

No file selected
Datasets used for X-DILiver development: Download

Note: Chemical compounds from the CDDI database have been excluded from the dataset in accordance with licensing restrictions.

Citation: Lee, J., Yu, MS., Lee, Y., Nam, H.J., Oh, KS., Jang, J., Lee, HM. & Na, D. X-DILiver: an ensemble learning framework for predicting drug-induced liver injury. Sci. Rep. (2026). https://doi.org/10.1038/s41598-026-59299-9
Bug reports: blisszen@cau.ac.kr
jg0715@cau.ac.kr