CaDRReS-Sc

CaDRReS-Sc predicts clone-specific therapeutic vulnerabilities from single-cell RNA sequencing (scRNA-seq) transcriptomic signatures to account for intra-tumor heterogeneity in precision oncology.

Key Features:

• Single-Cell RNA Sequencing Integration: Leverages single-cell RNA sequencing (scRNA-seq) data to analyze transcriptomic heterogeneity within tumors at the clone level.
• Recommender System Framework: Uses a recommender system architecture to process large-scale cancer omics datasets and identify heterogeneous gene-expression signatures correlated with drug response, with reported predictive accuracy around 80%.
• Monotherapy and Combinatorial Predictions: Validated on patient-proximal cell lines, predicting monotherapy responses with Pearson correlation coefficient r > 0.6 and improving combinatorial therapy prediction performance by >10%.
• In Silico Drug Repurposing: Applies in silico screening across scRNA-seq compendiums to prioritize existing drugs targeting clone-specific vulnerabilities.

Scientific Applications:

• Clone-specific vulnerability identification: Identifies clone-specific therapeutic vulnerabilities within heterogeneous tumor cell populations using scRNA-seq-derived signatures.
• Precision oncology treatment prediction: Predicts drug responses to support personalized targeting of tumor clones using predicted monotherapy and combination sensitivities.
• Drug repurposing screens: Enables in silico drug-repurposing screens across scRNA-seq compendiums to prioritize candidate therapeutics.

Methodology:

Processes single-cell RNA sequencing (scRNA-seq) and large-scale cancer omics datasets using a recommender system architecture to identify gene-expression signatures correlated with drug response; validation used patient-proximal cell lines with Pearson correlation and combinatorial improvement metrics.