Phenotypic deconvolution in heterogeneous cancer cell populations using drug-screening data

Area of Science:

• Oncology
• Computational Biology
• Systems Biology

Background:

• Tumor heterogeneity, characterized by drug-tolerant or resistant subpopulations, is a major cause of cancer treatment failure and recurrence.
• Traditional genomic deconvolution methods struggle to accurately map genomic variations to functional drug responses.
• Understanding phenotypic heterogeneity is crucial for developing effective cancer therapies.

Purpose of the Study:

• To develop a novel statistical framework for profiling phenotypic heterogeneity in bulk tumor samples.
• To identify distinct tumor subpopulations with differential drug responses and estimate their frequencies and sensitivities.
• To enable individualized predictions of tumor growth under various therapeutic strategies.

Main Methods:

• Leveraging mechanistic population modeling to create a statistical framework.
• Applying the PhenoPop method to analyze standard drug-screen data from bulk tumor samples.
• Validating the approach using synthetic cell populations, mixed cell-line experiments, and patient samples.

Main Results:

• PhenoPop reliably identifies tumor subpopulations with varying drug sensitivities.
• The method accurately estimates the frequencies of these subpopulations within bulk samples.
• Individualized predictions of tumor growth under different therapies were demonstrated.

Conclusions:

• PhenoPop offers a powerful tool for dissecting phenotypic heterogeneity in cancer.
• This methodology can guide the development of personalized cancer treatment strategies.
• The approach has broader applicability to deconvolution problems in diverse biological systems.