INTEGRATED DATA ANALYSIS FOR HIGH THROUGHPUT BIOLOGY
June 13 - 26, 2007
Application Deadline: March 15, 2007

Instructors:
Harmen Bussemaker, Columbia University
Vincent Carey, Harvard University
Partha Mitra, Cold Spring Harbor Laboratory
Mark Reimers, National Cancer Institute

High-throughput biology, epitomized by the ubiquitous DNA microarray, is rapidly generating enormous observation sets. Biologists seeking to make sense of this growing body of data need to have a firm grasp of statistical methodology. This course is designed to build competence in quantitative methods for the analysis of high-throughput molecular biology data, from which meaningful inferences about biological processes can be drawn.

- Review of multivariate statistics
- R mini-tutorial
- Expression and other microarrays - experimental design, scanning and image analysis, quality control, normalization and probe-level analysis for spotted arrays or prefabricated chips, exploratory analysis, tests of significance and multiple testing, using R and Bioconductor
- Discrimination and classification of samples
- Identifying general regulation themes (e.g. Gene Ontology categories) in gene lists by statistical means
- Promoter analysis in yeast using CHIP and expression data
- Identifying regulatory polymorphism using SNP and expression data
- Characterizing the effect of DNA amplifications and deletions on gene expression in cancer using CGH and expression data on the same samples

Speakers in last year's course included:
Keith Baggerly, M.D. Anderson Cancer Centre
Vivian Cheung, University of Pennsylvania
Aedin Culhane, Dana Farber Cancer Institute/Harvard
Bruce Futcher, SUNY Stony Brook
Audrey Gasch, University of Wisconsin-Madison
Rafael Irizarry, Johns Hopkins Bloomberg School of Public Health
Vishy Iyer, University of Texas at Austin
Ari Melnick, Albert Einstein College of Medicine
Stefano Monti, Whitehead Institute/MIT
Terry Speed, University of California, Berkeley
Richard Spielman, University of Pennsylvania
John Weinstein, National Cancer Institute
Richard Young, Whitehead Institute/MIT
Julia Zeitlinger, MIT

The first week of the course will concentrate on analysis of specific types of microarray data (expression, Affymetrix, CGH, CHIP-chip, and SNP arrays), and proteomics. The second week will explore biological problems involving the integration of several types of high-throughput data. Data sets will be drawn from yeast, human polymorphisms, and cancer biology.

Students are expected to take some time before the course to become familiar with the R statistical programming environment. See:
http://www.r-project.org http://www.biostat.harvard.edu/~carey/cdataSetup.html

This course is supported with funds provided by the National Cancer Institute