Skip to Main Content
Copy number aberrations (CNAs) are frequently found in cancer genomes and believed to be tumorigenic. Unfortunately, CNAs often occur in wide regions of the cancer genome that harbor a large number of genes, making it a challenge to identify the candidate cancer drivers. Further, subtypes of cancer may be characterized with distinct CNA patterns and hence have different drivers. Here, we report a systematic method to automate the identification of candidate drivers in cancer subtypes. Specifically, we propose an iterative approach that alternates between kernel based gene expression clustering and gene signature selection. We applied the method to datasets of the pediatric cancer medulloblastoma (MB). A cross-dataset comparison indicates the robustness of our subtyping method. Based on the identified subtypes, we developed a PCA based approach for subtype-specific identification of cancer drivers. The top-ranked driver candidates are found to be enriched with known pathways in certain subtypes of MB. This might reveal new understandings for these subtypes.