New insight for pharmacogenomics studies from the transcriptional analysis of two large-scale cancer cell line panels

Area of Science:

• Genomics
• Pharmacogenomics
• Cancer Biology

Background:

• High attrition rates in anticancer drug development necessitate innovative approaches.
• Drug repositioning offers a strategy to identify new uses for existing drugs.
• Optimizing large-scale pharmacogenomics datasets requires advanced analytical methods.

Purpose of the Study:

• To develop a novel molecular classification of cancer cell lines based on transcriptomic profiles.
• To improve the homogeneity of drug sensitivity within classified cell line groups.
• To identify robust associations between molecular clusters and drug response for drug repositioning.

Main Methods:

• Analysis of pharmacogenomics data from the Genomics of Drug Sensitivity in Cancer and Cancer Cell Line Encyclopedia studies.
• Focus on common cell lines (n=471) and common drugs (n=15).
• Development of a novel classification using transcriptomic profiles and a network-driven gene selection process.

Main Results:

• The proposed molecular classification demonstrated greater drug sensitivity homogeneity compared to tissue-of-origin grouping.
• Significant and robust associations were identified between cell line clusters and drug response across datasets.
• The method's relevance was confirmed using external datasets and varied sensitivity metrics.

Conclusions:

• A robust molecular classification of cancer cell lines was established.
• This classification can enhance the identification of new therapeutic indications for existing drugs.
• The findings support the utility of network-driven transcriptomic profiling for drug repositioning in oncology.