Cancer progression: Model, therapy & extraction
Item
-
Title
-
Cancer progression: Model, therapy & extraction
-
Identifier
-
d_2009_2013:387b6c7375a9:12028
-
identifier
-
12699
-
Creator
-
Olde Loohuis, Loes,
-
Contributor
-
Rohit Parikh | Bud Mishra
-
Date
-
2013
-
Language
-
English
-
Publisher
-
City University of New York.
-
Subject
-
Computer science | Bioinformatics
-
Abstract
-
In this thesis we develop Cancer Hybrid Automata (CHAs), a modeling framework based on hybrid automata, to model the progression of cancers through discrete phenotypes. Both transition timing between states as well as the effect of drugs and clinical tests are parameters in the framework, thus allowing for the formalization of temporal statements about the progression as well as timed therapies. Using a game theoretical formulation of the problem we show how existing controller synthesis algorithms can be generalized to CHA models, so that (timed) therapies can be automatically generated. In the second part of this thesis we connect this formal framework to cancer patient data, focusing on copy number variation (CNV) data. The underlying process generating CNV segments is generally assumed to be memory-less, giving rise to an exponential distribution of segment lengths. We provide evidence from TCGA data suggesting that this generative model is too simplistic, and that segment lengths follow a power-law distribution instead. We show how an existing statistical method for detecting genetic regions of relevance for cancer can be improved through more accurate (power-law) null models. Finally, we develop an algorithm to extract CHA-like progression models from cross-sectional patient data. Based on a performance comparison on synthetic data, we show that our algorithm, which is based on a notion of probabilistic causality, outperforms an existing extraction method.
-
Type
-
dissertation
-
Source
-
2009_2013.csv
-
degree
-
Ph.D.
-
Program
-
Computer Science
