University of Southern California
Ray R. Irani Hall
Molecular and Computational Biology
Computational Biology Colloquium
Nancy Zhang
Stanford University
"Simultaneous Change-point Models with Applications to Cross-sample and Cross-platform Analysis of DNA Copy Number"
Abstract:
DNA copy number analysis involves the detection of chromosomal gains and losses using
high-density microarray platforms. Change-point methods have been applied successfully to
detecting signals in single data sequences derived from one biological sample. However, it is
common to have data sets involving hundreds to thousands of biological samples. How should
information be combined across samples to detect population level common polymorphisms?
Also, how should the samples be summarized to give a sparse signature of variation across
the cohort? It is also now common to have the same biological sample assayed using multiple
experimental platforms. For example, in the Cancer Genome Atlas project, each biological
sample is processed using Illumina, Affymetrix and Agilent chips. How should data be inte-
grated across platforms to achieve higher accuracy?
I will discuss the statistical issues underlying these problems and formulate a class of si-
multaneous change-point models for cross-sample and cross-platform data integration. These
models lead to interpretable scan statistics whose significance level can be theoretically an-
alyzed. I will also discuss model selection approaches for this class of models. The insights
gained from this study can be applied to integrative analysis of data from other types of
genome-wide profiling experiments, such as methylation or RNA expression.
Thursday, October 22, 2009
2:00 pm
RRI 101
