_{Wednesday, August 15}

_{Pre-Conference Short Course One (SC1)
8:30-12:00 *}

Optimizing Microarray Gene Expression Data: 
Insights, Experiences, and Results

Microarray experiments are costly - both in time and resources, making the careful design of experiments to generate useful gene expression data for diagnostics, target identification, screening, genotyping and other applications critical. Learn from these savvy, seasoned microarray experts as they share their insights and experiences in optimizing microarray gene expression data to obtain results. 

8:00 am Short Course One Registration

8:30 Moderator’s Remarks

8:35 Automated Target Preparation for Gene Expression Microarrays 
Frederic Raymond, M.S., Research Scientist, Functional Henomics, Microarray Facility, Nestle Research Center 
While technical variability in sample preparation and chip processing may influence microarray-derived gene expression results, standardized and automated procedures can generate reproducible and comparable results across experimental replicates, studies and even platforms. We have therefore equipped our laboratory with a liquid handling system and programmed it to automatically perform all steps of the microarray target preparation, i.e. all procedures upstream the on-chip hybridization. Our fully integrated robotic platform can process up to 24 samples in parallel and reproducibly produces high-quality biotin-labeled cRNAs ready to be hybridized on one-colour oligonucleotide microarrays. We present technology development and applications.

9:05 Interpreting Microarray Experiments Using Functional Genomics
Nandini Raghavan, Ph.D., Principal Biostatistician, Non-Clinical Biostatistics, Johnson & Johnson Pharmaceutical Research & Development LLC 
The accurate interpretation of results from microarray experiments has proven to be a challenging exercise. Much of the research effort thus far has focused on gene-specific analyses, and in addressing the ensuing multiple comparisons issue that arises when studying thousands of genes simultaneously. These approaches ignore relationships among genes, which can be useful for understanding the biological mechanisms underlying the observed expression changes. The evolution of gene annotation databases has now made it feasible to incorporate biological information about genes into the analysis. We present a structured approach to incorporating gene function and gene pathway information into the analysis of gene expression data. We present case studies which illustrate that analyses based on gene functions yield functionally more interpretable results than gene-specific analyses.

9:35 Tissue-Selective Genes and Their Roles in Drug Discovery
Wei Liu, Ph.D., Principal Scientist, Bioinformatics, Wyeth Research
With the existing vast amount of gene expression profiling data it is now possible to define genes that are selectively expressed in different tissues across the human body. The tissue-selective genes can provide insights to their functions and physiological roles in their selective tissues. An example will be given to illustrate the pathophysiological role of immune cells in chronic diseases as revealed by genes selectively expressed in various immune cells. 

10:05 Coffee Break

10:30 Too Much Data, but Little Inter-Changeability: A Lesson Learned from Mining Public Data on Tissue Specificity of Gene Expression
Dan Li, Ph.D., Senior Research Scientist, Informatics, Eli Lilly and Company
Our objective is to test the data comparability between SAGE and microarray technologies, through examining gene expression profiles under normal physiological states across a variety of tissues. Comparative analysis indicated that there are significant discrepancies in gene expressions based on the microarray and SAGE platforms. Our analysis has revealed that the discrepancy is not likely caused by the heterogeneity of tissues used in these technologies, or other spurious correlations resulting from microarray probe design, abundance of genes, or gene function. Further investigation has suggested the discrepancy can be partially explained by errors in the original assignment of SAGE tags to genes due to the evolution of sequence databases. In addition, sequence analysis has indicated that many SAGE tags and Affymetrix array probe sets are mapped to different splice variants or different sequence regions although they represent the same gene, which also contributes to the observed discrepancies between SAGE and array expression data.

11:00 Exploring Compound Activities through Dose Design and Pathway Analysis
Rui-Ru Ji, Ph.D., Senior Research Investigator, Applied Genomics, Bristol-Myers Squibb Co.
Expression profiling may be performed at different stages of drug discovery to assess compound potency, selectivity, on- and off-target activities, and effects due to chemotype. Such profiling experiments are often done by selecting a few doses that are marked as either ‘high’ or ‘low’ based on the compounds’ EC50 values in some other assay. We have devised a new strategy that exploits a dose-range design for the profiling. A novel scanning approach has been developed to identify genes that exhibit dose-response behavior and assign an EC50 for each response. In addition, pathway analyses are performed to reveal underlying biology across the dose range. This strategy is being employed to identify distinct activities within and between compounds.

11:30 FDA Perspective
Elizabeth Mansfield, Ph.D. Senior Staff Fellow, OIVD/CDRH/FDA 

PRE-CONFERENCE SHORT COURSE TWO* (SC2)

1:30 PM Short Course Two Registration 

2:00-5:00 Statistical Analysis & Experiment Design to Reduce Noise in the New Generation of RNA & DNA Microarrays
Thomas J. Downey, Jr., President & CEO, Partek, Inc.
Microarray data contains treatment and/or phenotype effects embedded in a sea of technical and biological noise. This workshop will demonstrate how to use experiment design and statistical analysis to reliably identify biological effects of interest while controlling and removing noise due to biological and technical nuisance effects. Attendees will learn how to employ completely randomized block designs and statistical analysis to isolate and remove batch effects due to processing batches, etc. clearly revealing the signals from the biological factors of interest. The techniques will be demonstrated using gene expression, copy number, exon, and ChIP-on-Chip regulation studies on tiling arrays. Attendees will learn how to apply and interpret statistical techniques such as analysis of variance (ANOVA), Hidden Markov Models (HMM) to identify differential gene expression, alternative splicing, copy number aberrations, and regions of protein/DNA binding. 

Who Should Attend 
Medical researchers who are or will be conducting studies using microarrays to interrogate RNA and DNA for relationships to disease and/or drug treatments.

* Separate Registration Required 

4:00-5:00 Early Conference Registration