Efficient algorithms to discover alterations with complementary functional association in cancer

Area of Science:

• Computational Biology
• Genomics
• Cancer Research

Background:

• Large-scale cancer studies generate vast datasets of somatic alterations.
• Mutual exclusivity of genetic alterations is a known pattern in cancer, indicating functional relationships.
• Quantitative target profiles offer additional data to refine the identification of cancer-related gene sets.

Purpose of the Study:

• To develop a method for finding groups of mutually exclusive genetic alterations associated with quantitative functional targets in cancer.
• To address the computational complexity of this problem and provide efficient algorithmic solutions.

Main Methods:

• Formulated the problem of finding mutually exclusive gene sets associated with functional targets.
• Developed and implemented two algorithms within the UNCOVER tool to solve this computationally hard problem.
• Evaluated UNCOVER's performance against state-of-the-art methods using analytic and experimental evidence.

Main Results:

• UNCOVER effectively identifies high-quality sets of mutually exclusive alterations associated with functional targets.
• The algorithms demonstrate superior performance compared to existing methods, even when using their evaluation metrics.
• UNCOVER significantly outperforms state-of-the-art methods in speed, enabling analysis of large-scale cancer cell line datasets.

Conclusions:

• UNCOVER provides an efficient and effective approach for discovering cancer-related gene sets with complementary functional associations.
• The tool successfully identified significant gene sets associated with functional targets in large datasets from Project Achilles and GDSC.
• The developed algorithms offer a substantial advancement in analyzing complex genomic data for cancer research.