Related Experiment Video
Updated: Jan 30, 2026
Author Spotlight: Establishing a Murine Non-Small Cell Lung Cancer Model for Developing Nanoformulations of Anticancer Drugs
Published on: May 10, 2024
Predict drug sensitivity of cancer cells with pathway activity inference
Xuewei Wang1, Zhifu Sun1, Michael T Zimmermann1,2
1Department of Health Sciences Research, Mayo Clinic, Rochester, MN, USA.
Background:
Predicting cellular responses to drugs has been a major challenge for personalized drug therapy regimen. Recent pharmacogenomic studies measured the sensitivities of heterogeneous cell lines to numerous drugs, and provided valuable data resources to develop and validate computational approaches for the prediction of drug responses. Most of current approaches predict drug sensitivity by building prediction models with individual genes, which suffer from low reproducibility due to biologic variability and difficulty to interpret biological relevance of novel gene-drug associations. As an alternative, pathway activity scores derived from gene expression could predict drug response of cancer cells.
Method:
In this study, pathway-based prediction models were built with four approaches inferring pathway activity in unsupervised manner, including competitive scoring approaches (DiffRank and GSVA) and self-contained scoring approaches (PLAGE and Z-score). These unsupervised pathway activity inference approaches were applied to predict drug responses of cancer cells using data from Cancer Cell Line Encyclopedia (CCLE).
Results:
Our analysis on all the 24 drugs from CCLE demonstrated that pathway-based models achieved better predictions for 14 out of the 24 drugs, while taking fewer features as inputs. Further investigation on indicated that pathway-based models indeed captured pathways involving drug-related genes (targets, transporters and metabolic enzymes) for majority of drugs, whereas gene-models failed to identify these drug-related genes, in most cases. Among the four approaches, competitive scoring (DiffRank and GSVA) provided more accurate predictions and captured more pathways involving drug-related genes than self-contained scoring (PLAGE and Z-Score). Detailed interpretation of top pathways from the top method (DiffRank) highlights the merit of pathway-based approaches to predict drug response by identifying pathways relevant to drug mechanisms.
Conclusion:
Taken together, pathway-based modeling with inferred pathway activity is a promising alternative to predict drug response, with the ability to easily interpret results and provide biological insights into the mechanisms of drug actions.
Insights
Pathway-based models accurately predict cancer drug responses by analyzing pathway activity, offering better interpretability and biological insights than gene-specific models. Competitive scoring methods like DiffRank and GSVA show superior performance in predicting drug sensitivity.
Area of Science:
- Computational biology
- Pharmacogenomics
- Cancer research
Background:
- Predicting cellular drug responses is crucial for personalized medicine.
- Current gene-based models lack reproducibility and interpretability.
- Pathway activity scores offer a promising alternative for drug response prediction.
Purpose of the Study:
- To develop and validate pathway-based models for predicting cancer drug response.
- To compare the performance of different unsupervised pathway activity inference approaches.
- To assess the biological relevance and interpretability of pathway-based predictions.
Main Methods:
- Utilized four unsupervised pathway activity inference approaches: DiffRank, GSVA, PLAGE, and Z-score.
- Applied these methods to predict drug responses using Cancer Cell Line Encyclopedia (CCLE) data.
- Compared pathway-based models against gene-based models for 24 drugs.
Main Results:
- Pathway-based models outperformed gene-based models in predicting drug responses for 14 out of 24 drugs.
- Competitive scoring approaches (DiffRank, GSVA) yielded more accurate predictions and identified more drug-related pathways.
- Pathway-based models successfully captured pathways involving drug targets, transporters, and metabolic enzymes, unlike gene-based models.
Conclusions:
- Pathway-based modeling using inferred pathway activity is a robust alternative for predicting drug response.
- This approach enhances interpretability and provides biological insights into drug mechanisms.
- Competitive scoring methods are recommended for pathway-based drug response prediction.
Related Concept Videos
01:13Sensitivity, Specificity, and Predicted Value
Sensitivity is the...
01:18Antiepileptic Drugs: GABAergic Pathway Potentiators
The key GABA pathway potentiators used in epilepsy management are as follows.
Benzodiazepines are a well-known class of drugs used for...
02:27Predicting Molecular Geometry
01:27C4 Pathway and CAM
C4 Pathway
The C4 pathway is used by plants such as...
01:15Theory of Attribution I: Correspondent Inference Theory
01:03Prediction Intervals
However, the point estimate is most likely not the exact value of the population parameter, but close to it. After calculating point estimates, we construct interval estimates, called confidence intervals or prediction intervals. This prediction interval comprises a range of values unlike the point estimate and is a better predictor of the observed sample value, y.