Deciphering genomic alterations in colorectal cancer through transcriptional subtype-based network analysis

• Published in: PloS one Vol. 8; no. 11; p. e79282
• Main Authors: Zhu, Jing, Wang, Jing, Shi, Zhiao, Franklin, Jeffrey L, Deane, Natasha G, Coffey, Robert J, Beauchamp, R Daniel, Zhang, Bing
• Format: Journal Article
• Language: English
• Published: United States Public Library of Science 15.11.2013; Public Library of Science (PLoS)
• Subjects: Alterations; Analysis; Animals; Cancer; Cancer genetics; Cancer therapies; Clustering; Cohort Studies; Colon; Colorectal cancer; Colorectal carcinoma; Colorectal Neoplasms - genetics; Colorectal Neoplasms - metabolism; Copy number; Datasets; Developmental biology; Embryogenesis; Embryonic development; Embryonic growth stage; Female; Gene expression; Gene Expression Regulation, Neoplastic - genetics; Genetic aspects; Genome, Human; Genomes; Genomics; Health aspects; Humans; Informatics; Male; Medical prognosis; Medical research; Medicine; Mesenchyme; Mice; Mutation; Neoplasm Proteins - biosynthesis; Neoplasm Proteins - genetics; Network analysis; Networks; Neural Networks, Computer; Random walk; Signal transduction; Signaling; Statistical analysis; Statistical methods; Transcription; Transcription (Genetics); Vascular endothelial growth factor; Wnt protein; Wnt Signaling Pathway - genetics; United States > US; Tennessee; Nashville Tennessee
• ISSN: 1932-6203; 1932-6203
• DOI: 10.1371/journal.pone.0079282

Both transcriptional subtype and signaling network analyses have proved useful in cancer genomics research. However, these two approaches are usually applied in isolation in existing studies. We reason that deciphering genomic alterations based on cancer transcriptional subtypes may help reveal subtype-specific driver networks and provide insights for the development of personalized therapeutic strategies. In this study, we defined transcriptional subtypes for colorectal cancer (CRC) and identified driver networks/pathways for each subtype. Applying consensus clustering to a patient cohort with 1173 samples identified three transcriptional subtypes, which were validated in an independent cohort with 485 samples. The three subtypes were characterized by different transcriptional programs related to normal adult colon, early colon embryonic development, and epithelial mesenchymal transition, respectively. They also showed statistically different clinical outcomes. For each subtype, we mapped somatic mutation and copy number variation data onto an integrated signaling network and identified subtype-specific driver networks using a random walk-based strategy. We found that genomic alterations in the Wnt signaling pathway were common among all three subtypes; however, unique combinations of pathway alterations including Wnt, VEGF and Notch drove distinct molecular and clinical phenotypes in different CRC subtypes. Our results provide a coherent and integrated picture of human CRC that links genomic alterations to molecular and clinical consequences, and which provides insights for the development of personalized therapeutic strategies for different CRC subtypes.