Related Experiment Video
Updated: Jun 28, 2026
Non-Destructive Evaluation of Regional Cell Density Within Tumor Aggregates Following Drug Treatment
Published on: June 21, 2022
scRADAR: Dissecting intratumoral drug response heterogeneity at single-cell resolution via mechanism-guided prototype
Ren Qi1, Wenjie Teng1, Xin Yang1
1School of Computer Science and Technology, Beijing Institute of Technology, Beijing, China.
Abstract:
Precision oncology requires resolving intratumoral heterogeneity to identify drug-resistant cell states associated with treatment failure and relapse. Although single-cell RNA sequencing enables characterization of heterogeneous resistance-associated states, single-cell drug-response phenotype prediction remains challenging because of sparsity, noise, class imbalance, and limited mechanistic interpretability. Here, we present scRADAR (Response Analysis via Drug-Aware Routing), a mechanism-guided prototype routing framework for predicting and interpreting drug-response phenotypes at single-cell resolution. Rather than relying on cell-line-anchored transfer learning, scRADAR learns directly from labeled single-cell cohorts. The framework integrates metabolic and signaling pathway activities to form a dual-view cellular representation, conditions pathway embeddings on drug mechanisms through feature-wise linear modulation, and uses sparse prototype routing to decompose predictions into interpretable response archetypes. Across nine independent cohorts, scRADAR showed strong predictive performance and consistent cross-cohort behavior, particularly under imbalanced settings. Post hoc attribution analyses highlighted candidate TGF-β-associated epithelial-to-mesenchymal transition signatures in Erlotinib-associated Resistant-labeled states and cytoskeletal/metabolic response-associated signatures in BET-inhibitor-associated Resistant-labeled states. These results suggest that scRADAR provides an interpretable framework for single-cell drug-response phenotype prediction and for generating hypotheses about resistance-associated programs from heterogeneous tumor transcriptomes.
Insights
We developed scRADAR, a new computational framework to predict single-cell drug responses. This tool helps identify drug-resistant cancer cells by analyzing tumor heterogeneity, improving precision oncology.
Area of Science:
- Computational Biology
- Genomics
- Precision Medicine
Background:
- Precision oncology aims to overcome treatment failure by understanding tumor heterogeneity.
- Single-cell RNA sequencing (scRNA-seq) can identify drug-resistant cell states, but predicting drug response at this resolution is difficult due to data challenges.
- Existing methods often struggle with interpretability and direct learning from patient data.
Purpose of the Study:
- To introduce scRADAR (Response Analysis via Drug-Aware Routing), a novel framework for predicting and interpreting drug-response phenotypes at the single-cell level.
- To address limitations in current single-cell drug-response prediction, including sparsity, noise, class imbalance, and lack of mechanistic interpretability.
- To enable hypothesis generation regarding cancer drug resistance mechanisms directly from heterogeneous tumor transcriptomes.
Main Methods:
- scRADAR employs a mechanism-guided prototype routing framework that learns directly from labeled single-cell cohorts.
- It integrates metabolic and signaling pathway activities for a dual-view cellular representation.
- Pathway embeddings are conditioned on drug mechanisms using feature-wise linear modulation, and sparse prototype routing decomposes predictions into interpretable archetypes.
Main Results:
- scRADAR demonstrated strong predictive performance across nine independent cohorts, showing robust behavior even with imbalanced datasets.
- The framework successfully predicted drug-response phenotypes at single-cell resolution.
- Post hoc analyses identified potential resistance mechanisms, including TGF-β-associated EMT signatures for Erlotinib resistance and cytoskeletal/metabolic signatures for BET-inhibitor resistance.
Conclusions:
- scRADAR offers an interpretable framework for single-cell drug-response phenotype prediction in precision oncology.
- The tool can generate testable hypotheses about the molecular programs underlying drug resistance.
- This approach advances the analysis of heterogeneous tumor transcriptomes for personalized cancer treatment.
