Schedule:	Spring 2004
Schedule:	2:50-5:30pm, Mon
Place:	130B North Building
Instructors:	Joe Nevins, Department of Molecular Genetics & Microbiology Mike West, Duke Statistics

Prerequisites: Expertise and facility in molecular biology and genetics at the level of BGT 203 & BGT 206 . Expertise and facilty in statistical theory, methods and computation at the level of BGT 200 & STA 244 -- students must be proficient in statistics and computing at this level, including programming with applied computing environments (Matlab, R/S-Plus) as well as in low level languages (C/C++ or Fortran).

Students will program either on the Duke Statistics system or on personal computers, the latter involving download of the free R software and a number of related packages, an Matlab installation under Duke site licenses (via the OIT web site) .

Coverage/Topics:

• Biological phenotypes
• Gene expression: principles, biology and biochemistry
  - gene structure
  - gene regulation
  - transcription
  - gene pathways
  - concept of steady state

  Gene expression: measurement, technology and applications

  - RNA isolation, hybridization

  - Northern analysis

  - RT-PCR

  DNA microarrays - technology, data and analytics

  - Oligonucleotide technologies (Affymetrix)

  - Printed cDNA and oligo arrays

  - Informatics aspects: data processing and scanning, data-basing, data quality and evaluation

  Statistics and computation

  - graphical and numerical exploratory data analysis

  - multivariate analysis, classification and discrimination

  - prediction and regression analysis methods

  - factor analysis and data representation & reduction

  - data processing, imaging, normalization & other aspects of low-level analysis

  Experimental generation of microarray data (wet lab component)

  - isolation of RNA from cell cultures

  - analysis of RNA quality

  Applications in

  - phenotyping: understanding genes implicated in clinical and physiological outcomes

  - various applications in cancer phenotyping and pathway studies

  - pattern analysis in cell cycle regulations, and other areas

  - pathway exploration

  Statistical modelling

  - EM, search and MCMC methods for fitting statistical models

  - Statistical computing for non-linear and regularised regression

  - Survival analysis: e.g., predicting recurrence in cancer based on gene expression profiles

  - Non-linear association, retrospective studies

  - Graphical models for visualisation and exploration

  Information systems and development of integrated tools

  - DIG information systems

  - Gene annotations and public data bases

  - Annotation access tools and connecting with statistical visualization tools