Related Experiment Video
Updated: Aug 3, 2026
Multiparametric Tumor Organoid Drug Screening Using Widefield Live-Cell Imaging for Bulk and Single-Organoid Analysis
Published on: December 23, 2022
Carcinogenicity prediction via multi-task learning of cross-organ representations with attention mechanisms
Yunju Song1, Hwan Choi2, Sunyong Yoo1,3
1Department of Intelligent Electronics and Computer Engineering, Chonnam National University, 77, Yongbong-ro, Buk-gu, Gwangju, 61186, Republic of Korea.
Abstract:
Cancer is caused by the uncontrolled growth and division of abnormal cells. In industrialized societies, chemical exposure is a leading cause of cancer. Since certain compounds induce cancer by damaging genes or affecting cellular metabolism, studying carcinogens is essential. However, previous studies used separate models for each organ and failed to capture carcinogenic features shared across organs, limiting generalization. Thus, this study developed a multi-task learning framework to predict organ-specific carcinogenicity in the liver, lung, stomach, and breast. This framework consisted of a shared layer and task-specific layers. The shared layer contains a graph attention network layer to make atom-level representations, along with parallel fully connected layers designed for each task combination. The resulting shared representations are passed to task-specific layers to predict organ-specific carcinogenicity. The training process followed stepwise learning, whereby the model was first trained using partially labeled data to capture cross-organ representations and determine initial weights. In the second step, fully labeled data for all organs were used for final training. The proposed multi-task model achieved superior performance in the liver, lung, and stomach tasks. Notably, it recorded the highest area under the receiver operating characteristic curve in the stomach task (0.7636), outperforming the single-task model (0.7055) and all comparative models (0.5527-0.7418). The highest area under the precision-recall curve was observed in the liver task (0.9646), surpassing the single-task model (0.9505) and all comparative models (0.9373-0.9621). We further analyzed molecules with high predicted carcinogenicity and identified critical substructures using an attention mechanism. This research can contribute to predicting organ-specific carcinogenicity of candidate chemicals in the early stages of drug development.
Insights
This study introduces a novel multi-task learning framework to predict chemical carcinogenicity across multiple organs. The model effectively captures shared carcinogenic features, improving prediction accuracy for liver, lung, and stomach cancers.
Area of Science:
- Toxicology and Cheminformatics
- Computational Biology
- Machine Learning in Drug Discovery
Background:
- Chemical exposure is a significant cause of cancer in industrialized societies.
- Previous models for carcinogenicity prediction were organ-specific, limiting generalization of findings.
- A unified approach is needed to identify shared carcinogenic mechanisms across different organs.
Purpose of the Study:
- To develop and evaluate a multi-task learning framework for predicting organ-specific chemical carcinogenicity.
- To improve the accuracy and generalization of carcinogenicity predictions by leveraging shared features across organs.
- To identify critical molecular substructures associated with carcinogenicity using attention mechanisms.
Main Methods:
- A multi-task learning framework integrating a graph attention network for atom-level representations and task-specific layers.
- Stepwise learning strategy involving initial training on partially labeled data followed by final training on fully labeled data.
- Evaluation using Area Under the Receiver Operating Characteristic Curve (AUC-ROC) and Area Under the Precision-Recall Curve (AUC-PR).
Main Results:
- The multi-task model demonstrated superior performance for liver, lung, and stomach carcinogenicity prediction compared to single-task models.
- Achieved the highest AUC-ROC in the stomach task (0.7636) and highest AUC-PR in the liver task (0.9646).
- Attention mechanism analysis identified critical molecular substructures contributing to predicted carcinogenicity.
Conclusions:
- The proposed multi-task learning framework effectively predicts organ-specific carcinogenicity by learning shared representations.
- This approach enhances prediction accuracy and offers insights into cross-organ carcinogenic mechanisms.
- The model has potential applications in early-stage drug development for chemical safety assessment.