Drug Repositioning for Cancer Therapy Based on Large-Scale Drug-Induced Transcriptional Signatures

Area of Science:

• Computational biology
• Genomics
• Drug discovery

Background:

• Drug repositioning (DR) offers a faster route to new cancer therapies.
• Existing methods often rely on drug structure or known targets, limiting discovery.
• Large-scale drug-induced gene expression datasets provide rich information on drug effects.

Purpose of the Study:

• To develop and validate an in silico chemical genomics approach for predicting drug repositioning candidates for glioblastoma, lung cancer, and breast cancer.
• To leverage drug-induced expression profiles for more effective DR candidate identification compared to traditional methods.

Main Methods:

• Utilized a dataset of ~20,000 drug-induced expression profiles across multiple cancer cell lines.
• Employed an in silico chemical genomics strategy to analyze transcriptional perturbation and drug mode-of-action.
• Validated the predictive power of expression signatures against high-throughput screening (HTS) datasets.

Main Results:

• Drug-induced expression profiles proved more effective for DR prediction than drug structure or known targets.
• Identified 14 high-scoring DR candidates, with eight showing significant anti-proliferative activity against glioblastoma.
• The DR score strongly correlated with experimental results, achieving a ~20-fold enrichment for top candidates.

Conclusions:

• The developed in silico method accurately predicts drug repositioning candidates for cancer.
• Identified specific drugs like ivermectin and trifluridine as promising glioblastoma treatments.
• Discovered common transcriptome-level mechanisms underlying the anti-proliferative activity of repositioned drugs.