Related Experiment Video
Updated: Jan 31, 2026
Patient-Derived Tumor Explants As a "Live" Preclinical Platform for Predicting Drug Resistance in Patients
Published on: February 7, 2021
Predicting tumor cell line response to drug pairs with deep learning
Fangfang Xia1,2, Maulik Shukla3, Thomas Brettin3
1Computing, Environment and Life Sciences, Argonne National Laboratory, Lemont, IL, USA. fangfang@anl.gov.
Background:
The National Cancer Institute drug pair screening effort against 60 well-characterized human tumor cell lines (NCI-60) presents an unprecedented resource for modeling combinational drug activity.
Results:
We present a computational model for predicting cell line response to a subset of drug pairs in the NCI-ALMANAC database. Based on residual neural networks for encoding features as well as predicting tumor growth, our model explains 94% of the response variance. While our best result is achieved with a combination of molecular feature types (gene expression, microRNA and proteome), we show that most of the predictive power comes from drug descriptors. To further demonstrate value in detecting anticancer therapy, we rank the drug pairs for each cell line based on model predicted combination effect and recover 80% of the top pairs with enhanced activity.
Conclusions:
We present promising results in applying deep learning to predicting combinational drug response. Our feature analysis indicates screening data involving more cell lines are needed for the models to make better use of molecular features.
Insights
This study developed a deep learning model to predict how cancer cell lines respond to drug combinations, achieving 94% accuracy. The model prioritizes drug properties over molecular features for effective anticancer therapy predictions.
Area of Science:
- Computational biology
- Pharmacogenomics
- Artificial intelligence in medicine
Background:
- The National Cancer Institute-60 (NCI-60) cell line panel provides a valuable resource for studying drug interactions.
- Understanding combinational drug activity is crucial for developing effective cancer therapies.
Purpose of the Study:
- To develop a computational model for predicting drug pair response in cancer cell lines.
- To leverage deep learning for modeling combinational drug effects.
Main Methods:
- Utilized residual neural networks to encode features and predict tumor growth.
- Integrated molecular data (gene expression, microRNA, proteome) and drug descriptors.
- Applied the model to the NCI-ALMANAC database.
Main Results:
- The model explained 94% of the response variance in cell line drug pair screening.
- Drug descriptors were found to be the primary drivers of predictive power.
- Successfully identified 80% of top-performing drug pairs with enhanced anticancer activity.
Conclusions:
- Deep learning shows promise for predicting combinational drug response.
- Future models could benefit from larger screening datasets to better utilize molecular features.
More Related Videos
10:46Rapid in vivo Drug Response Prediction Using Leukemia Cell Grafts in Zebrafish Embryos
Published on: May 23, 2025
04:09Predicting Treatment Response to Image-Guided Therapies Using Machine Learning: An Example for Trans-Arterial Treatment of Hepatocellular Carcinoma
Published on: October 10, 2018
Related Concept Videos
02:27Predicting Molecular Geometry
04:01VSEPR Theory and the Effect of Lone Pairs
01:21Factors Affecting Drug Response: Overview
02:27DNA Base Pairing
02:27DNA Base Pairing
01:17Sign Test for Matched Pairs
To conduct the sign test, we first calculate the differences in...