TL
Tzong-Yi Lee
Author with expertise in Real-Time Polymerase Chain Reaction
Achievements
This user has not unlocked any achievements yet.
Key Stats
Upvotes received:
0
Publications:
3
(33% Open Access)
Cited by:
0
h-index:
3
/
i10-index:
1
Reputation
Biology
< 1%
Chemistry
< 1%
Economics
< 1%
Show more
How is this calculated?
Publications
0

StackDILI: Enhancing Drug-Induced Liver Injury Prediction through Stacking Strategy with Effective Molecular Representations

Jia-Hui Guan et al.Jan 9, 2025
Drug-induced liver injury (DILI) is a major challenge in drug development, often leading to clinical trial failures and market withdrawals due to liver toxicity. This study presents StackDILI, a computational framework designed to accelerate toxicity assessment by predicting DILI risk. StackDILI integrates multiple molecular descriptors to extract structural and physicochemical features, including the constitution, pharmacophore, MACCS, and E-state descriptors. Additionally, a genetic algorithm is employed for feature selection and optimization, ensuring that the most relevant features are used. These optimized features are processed through a stacking ensemble model comprising multiple tree-based machine learning models, improving prediction accuracy and interpretability. Notably, StackDILI demonstrates a strong performance on the DILIrank test set and maintains robustness across cross-validation. Moreover, interpretability analysis reveals key molecular features associated with DILI risks, providing valuable insights into toxicity prediction. To further improve accessibility, a user-friendly web interface is developed, allowing users to input SMILES strings and receive rapid predictions with ease. The StackDILI model provides a powerful tool for efficient DILI assessment, supporting safer drug development. The web interface is accessible at https://awi.cuhk.edu.cn/biosequence/StackDILI/.