\documentclass[11pt]{article}
\setlength{\oddsidemargin}{0.0truein}
\setlength{\evensidemargin}{0.0truein}
\setlength{\textwidth}{6.5truein}
\setlength{\topmargin}{0.0truein}
\setlength{\textheight}{9.0truein}
\setlength{\headsep}{0.0truein}
\setlength{\headheight}{0.0truein}
\setlength{\topskip}{10.0pt}
\usepackage{url}
\begin{document}

\begin{center}
  \textbf{\textsc{STANFORD UNIVERSITY}}\\[5pt]
  \textbf{\textsc{DEPARTMENT OF STATISTICS}}\\[5pt]
  \Large{\textbf\textsc{{DEPARTMENTAL SEMINAR}}}
\end{center}
\begin{center}
  4:15 p.m., Tuesday, June 14, 2005\\
  Sequoia Hall Room 200\\
  (Cookies at 3:45 in 1st Floor Lounge)
\end{center}
\begin{center}
  \textsl{Marc Coram}\\
  University of Chicago\\
  \textsl{Pei Wang}\\
  Fred Hutchinson Cancer Research Center\\
  \textsl{Hua Tang}\\
  Fred Hutchinson Cancer Research Center\\
\end{center}
\begin{center}
  \textbf{Statistical Challenges in Comparative Proteomics}
\end{center}
Proteomics research has an enormous clinical and scientific potential. However,
statistical challenges must be overcome before the full potential can be
realized. In this talk, we analyze mass spectrometry data on human blood serum.
The samples are chemically treated, separated by liquid chromatography, and
analyzed via electrospray ionization mass spectroscopy (LC/ESI-MS). We will
describe the statistical methodologies we have developed to analyze these data.
Major challenges include: (1) detecting peptide signatures, (2) quantifying
peptide intensity, (3) comparing proteomic profiles from run to run. These
challenges require scanning each vast and noisy image for the chains of peaks
that mark peptide locations. There are several complications. First, two types
of noise are present: instrumental noise and chemical noise. Second, sample
preparation can substantially distort the signal and noise intensities from run
to run. Third, while the mass/charge aspect of the data is quite accurate and
reproducible, the chromatographic aspect varies both globally, across each run,
and locally, from peptide to peptide. We describe some current approaches to
these challenges, and our work on a more unified approach.
\end{document}