Understanding the dynamical evolution of cancer, with the final goal of developing effective techniques for diagnosis, prediction and treatment is one of the main challenges of modern biosciences. In this paper we approach the temporal ordering reconstruction problem, which refers to the temporal sorting of a collection of static biological data. The solution of this problem may help in better understanding the key principles and properties of the disease progression. By using a previously proposed technique for extracting temporal progressions from cross-sectional cancer gene expression data, we develop a novel methodology to be applied to static cross-sectional copy number alterations, and we test it on patients diagnosed with colorectal cancer at different stages. To capture distinct aspects of this complex phenomenon, we define several measures of chromosomal alterations and filters targeting significant portions of chromosomes. Results obtained with various measures and filters highlight the best setting for the problem, the most relevant chromosomal alterations and emphasize the influence that copy number alterations hitting key genes may have on the development of the disease.
|Number of pages||25|
|Journal||International Journal of Data Mining and Bioinformatics|
|Publication status||Published - Jan 2016|