Ordering copy number alteration data to analyze colorectal cancer progression

Cancer is a very complex disease and understanding its dynamics and evolution is one of the challenges of modern biosciences. As most available data on cancer is static, extracting dynamic information about its progression from “static” biological data would have a major significance. We are approaching the Temporal Ordering Reconstruction (TOR) problem, that is the sorting of a collection of multi-dimensional biological data to reflect an accurate temporal progression of the target disease. The most general form of the TOR problem has been studied from many points of view. Firstly, the TOR problem, as defined above has been tackled mostly in two works, which use gene expression data as the “raw’” data in the samples (Gupta and Bar-Joseph, 2008; Magwene et al., 2003). Secondly, another series of works start by analyzing comparative genomic hybridization data to build a plausible tree of possible gene mutation events and continue towards a use of Bayesian models to assess pathways variations in a disease (Desper et al., 1999; Pathare et al., 2009; Gerstung et al., 2011; Beerenwinkel et al., 2005). Our work is more focused on a specific approach to the TOR problem, previously proposed by Gupta and Bar-Joseph (Gupta and Bar-Joseph, 2008), which has been shown to work for gene expression data and we develop a methodology which enables us to apply this technique on a Copy Number Alterations (CNAs) data set. We also aim to provide a building block in an analysis pipeline that can be used to look at temporal reconstruction problems that assume an already (partially) ordered dataset (Ramakrishnan, 2010; Antoniotti, 2010).