Detecting independent and recurrent copy number aberrations using interval graphs.

Area of Science:

• Genomics
• Computational Biology
• Cancer Research

Background:

• Somatic copy number aberrations (SCNAs) are common in cancer genomes but distinguishing functional events from random passengers is difficult.
• The variability in SCNA length and position complicates the identification of recurrent aberrations across samples.

Purpose of the Study:

• To develop a robust computational method for identifying independent and recurrent SCNAs.
• To overcome the challenge of overlapping aberrations in cancer genomes.

Main Methods:

• Introduced a combinatorial approach using maximal cliques in an interval graph to model SCNA overlaps.
• Developed a dynamic programming algorithm (RAIG) for efficient enumeration and selection of non-overlapping recurrent SCNA events.
• Optimized a well-defined objective function for selecting independent aberrations.

Main Results:

• RAIG efficiently identifies independent and recurrent SCNAs, outperforming existing methods on simulated and TCGA cancer data.
• The algorithm successfully separates overlapping aberrations into independent events.
• RAIG can detect rare, potentially functional aberrations obscured by larger passenger events.

Conclusions:

• RAIG provides a fast and effective method for identifying significant SCNAs in cancer genomics.
• The approach offers an advantage over heuristic methods by optimizing a defined objective.
• This facilitates the discovery of potentially driver SCNAs for cancer research.