psb-app

psb-app predicts patient-specific drug sensitivity by integrating patients, cancer cell lines, and drugs into a multilayer network to support precision oncology.

Key Features:

• Multilayer network-based approach: Constructs an interconnected network of patients, cell lines, and drugs to integrate high-throughput molecular profiling and drug sensitivity information.
• Patient-cell line matching: Links patient tumors to representative cancer cell lines using omics-profile-based network similarity measures.
• Personalized Imputed Drug Sensitivity Score (PIDS-Score): Imputes drug sensitivity scores for individual patients (PIDS-Scores) to indicate therapeutic potential for specific patient-drug pairs.
• Application to lung cancer with functional proteomics: Matches lung cancer patients to cell lines from 19 tissue types using functional proteomics profiles and computes PIDS-Scores for 251 drugs and experimental compounds.
• Clinical outcome association: Produces PIDS-Score-based predictions that are reported to be significantly associated with clinical outcomes in lung cancer patients.

Scientific Applications:

• Precision oncology: Supports selection of therapies tailored to individual tumor molecular profiles by predicting patient-specific drug responses.
• Research and development: Aids investigation of cancer biology and identification of molecular targets relevant to specific cancer subtypes.

Methodology:

Constructs a patient–cell line network using omics data, computes network-based similarity to determine best matches, imputes PIDS-Scores for drugs, applies functional proteomics for lung cancer matching across 19 tissue types and 251 drugs, and employs robust statistical measures and computational techniques to derive predictions.