Multi-Omics Alleviates the Limitations of Panel Sequencing for Cancer Drug Response Prediction

Area of Science:

• Oncology
• Genomics
• Pharmacogenomics

Background:

• Comprehensive genomic profiling using cancer gene panels aids cancer treatment selection.
• Genomic aberrations alone are often insufficient predictors of drug sensitivity.

Purpose of the Study:

• To evaluate the utility of incorporating transcriptome data with gene panel features for improving cancer drug response prediction.
• To determine if combining genomic and transcriptomic data enhances the accuracy of predicting treatment outcomes.

Main Methods:

• Utilized large-scale pharmacogenomics datasets encompassing cell lines, patient-derived xenografts, and ex vivo treated tumor specimens.
• Integrated gene panel mutation data with transcriptome expression profiles.
• Developed and validated predictive models for drug response.

Main Results:

• The addition of transcriptome data to gene panel features substantially improved drug response prediction performance.
• Combined genomic and transcriptomic analyses provided more robust predictions compared to genomic data alone.
• Demonstrated improved predictive accuracy across diverse cancer models.

Conclusions:

• Transcriptome data is a crucial addition to genomic profiling for accurate cancer drug sensitivity prediction.
• Integrating multi-omics data, specifically gene panels and transcriptome, offers a more powerful approach for precision oncology.
• This enhanced predictive capability can guide more effective and personalized cancer therapy selection.