Your Name: Chun Tang Your E-mail address: chuntang@cse.buffalo…

Your Name: Chun Tang
     Your E-mail address: chuntang@cse.buffalo.edu
     Your Dissertation Proposal Title: Mining Phenotypes and
Informative Genes Underlying Gene Expression Profiles
     Your Committee:
               Chair:     Dr. Aidong Zhang
               Members:   Dr. Jian Pei, Dr. Murali Ramanathan
(Department of Pharmaceutical Sciences)
     Your Dissertation Proposal Abstract:

Recently introduced DNA microarray technology permits rapid, large-
scale screening for patterns of gene expression and gives simultaneous,
semi-quantitative readouts on the level of expression of thousands of
genes for samples. The raw microarray data (images) can then be
transformed into gene expression matrices where usually a row in the
matrix represents a genes and a column represents a sample.
The numeric value in each cell characterizes the expression level of
the particular gene in a particular sample. Microarray technology has a
significant impact on the field of bioinformatics, requiring innovative
techniques to efficiently and effectively extract, analysis, and
visualize these fast growing data.

While most of the previous studies focus on clustering either genes or
samples, it is interesting to ask whether we can partition the complete
set of samples into exclusive groups (called phenotypes) and find a set
of informative genes that can manifest the phenotype structure
simultaneously. The mining of phenotypes and informative genes can
provide valuable information for biologists to understand the roles of
genes and the phenotypes of samples.

Most of the genes collected by microarray experiments may not
necessarily be task-specific. A small percentage of genes which
manifest the meaningful phenotype structure of the samples are buried
in large amount of noise. Uncertainty about which genes are relevant
makes it difficult to construct an informative gene space.
The number of genes and the number of samples are very different in a
typical gene expression matrix. Usually, we may have tens or hundreds
of samples but thousands or tens of thousands of genes.
Since the number of samples is usually limited, such data sets are very
sparse in high-dimensional genes space. Unsupervised phenotype
structure and informative gene discovery of such sparse high-
dimensional data sets presents an interesting but also very challenging
problem. No existing approaches can be effectively and efficiently used
to mine phenotypes and informative genes of such data sets.

In this proposal, we propose the new problem of mining phenotypes and
informative genes from gene expression data sets and a novel
unsupervised analyzing framework to detecting phenotypes and
informative genes underlying gene expression data sets.
A series of statistical measurements are proposed to measure the
quality of the mining results. These measurements delineate local
pattern qualities based on a partition of samples on a subset of genes
to coordinate between sample phenotype discovery and informative space
detection. Two interesting algorithms are developed: the heuristic
search and the mutual reinforcing adjustment method. Iterative pattern
adjustment strategies are presented to approach the optimal solution
which the pattern quality is maximized. The methods dynamically measure
and manipulate the relationship between samples and genes while
conducting an iterative adjustment of genes and samples to approximate
the informative genes and the phenotypes of the samples simultaneously.
We present an extensive performance study on both real-world data sets
and synthetic data sets. Our results strongly suggest that the two
proposed methods are effective and scalable. The mining results are
clearly better than the previous methods. They are ready for the real-
world applications. The mutual reinforcing adjustment method is in
general more scalable, more effective and with better quality of the
mining results.