Abstract 3758: A combined machine learning and bioinformatic analysis approach identifies biological pathways that predict clinical stage and survival outcome in neuroblastoma patients

Abstract

Abstract Neuroblastoma is an aggressive pediatric cancer that displays tremendous clinical and biological heterogeneity. Approximately 40% of patients present with metastatic disease and have poor outcome; new biomarkers that improve risk classification and prognostication are desperately needed. Applying a machine learning-based approach to analyzing transcriptome-wide signals in multiple populations of neuroblastoma patients spanning clinical stage we identified a combined set of 14 clinical and molecular variables that predicted survival outcome and tumor stage with over 87% accuracy. A systems biology-based analysis of the molecular variables from our predictive set identified 5 major networks with hubs defining 3 major ontological processes: apoptosis, nucleotide biosynthesis, and RNA metabolism. Citation Format: Alex Carlisle, Ivan Caceres, Sonali Mehta, Jay Schindler, Jonathan Sharma. A combined machine learning and bioinformatic analysis approach identifies biological pathways that predict clinical stage and survival outcome in neuroblastoma patients. [abstract]. In: Proceedings of the 106th Annual Meeting of the American Association for Cancer Research; 2015 Apr 18-22; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2015;75(15 Suppl):Abstract nr 3758. doi:10.1158/1538-7445.AM2015-3758