Tuesday, October 16
7:30 am Breakfast Technology Workshop
8:30 Back by Popular Demand
Java and Jive Break-out Discussion Groups
Time has been designated for facilitated, discussion groups with specific themes. This unique opportunity allows conference participants to focus on a topic and exchange ideas, information, experiences, and develop future collaborations.
9:30 Chairperson’s Remarks
9:35 Decipher Biologic Pathways and Determine the Total “Omic” Picture with an Integrated DAVID Knowledgebase
Richard Lempicki, Ph.D, Microarray Core Facility Head/Senior Scientist, Clinic Services Program, NIAID/NIH
Microarray analyses provides a wealth of information to help uncover complex biological processes, better understand disease pathogenesis, discover novel biomarker panels or design novel diagnostic methods, to mention a few. A key to success when embarking in such studies requires the effective use of sequentially applied procedures that reduce the false positive rate at each step as an investigator drills down to the most pertained information. The steps include: 1) good experimental design, 2) microarray quality control analysis, 3) application of proper microarray statistics and gene selection heuristics, 4) high quality biological interpretation, and 5) biological confirmation and validation using independent samples. I will illustrate the use of such a pipeline in a group of studies(n=180) aimed at better understanding peripheral blood mononuclear cell gene expression profiles during HIV infection in vivo, including: i) comparing results from the Affymetrix’s HuFL, U133A, U133 2.0 Plus, and Exon arrays, ii) biological interpretation using our freely available knowledgebase (DAVID: http://david.niaid.nih.gov), and iii) comparison of a high-throughput 30-plex bDNA/xMAP fluorescent bead method (QuantiGene Plex from Panomic, Inc.) to real time PCR for microarray confirmation and biomarker detection.
10:05 Integrated Microarray Data Analysis Revealed Pathological Pathways of Osteoarthritis
Tao Wei, Ph.D, Senior Research Scientist, Integrative Biology, Eli Lilly and Company
Osteoarthritis (OA) is one of the most frequent and symptomatic health problems for middle aged and older people (Wieland, Michaelis et al. 2005). However, the pathophysiological pathways are complex and poorly understood (Buckwalter and Martin 2006), which has hindered development of effective therapies. Microarray technology has been applied to study disease mechanisms of OA and OA models. It helps our understanding of disease mechanisms, thus drug development only if we could develop novel biological insights from a large amount of array data generated from different research labs. We applied integrated analysis of recently published microarray studies with our own array study in the articular cartilage of OA patients and animal models and identified several pathways implicated in cartilage degradation. The insights we developed not only demonstrate the power of the technology but also provide new classes of targets for drug development.
10:35 Technology Spotlight
10:50 Networking Coffee Break, Poster and Exhibit Viewing
11:30 Cross-site and Cross-platform Concordance Improved by Variance Stabilization
Pan Du, Ph.D, Research Associate, Robert H. Lurie Comprehensive Cancer Center, Northwestern University
Gene expression data is prone to variance from experiment to experiment, lab to lab, platform to platform. Improving the concordance between experiments, cross sites and platforms becomes very important for the result reproducibility and comparability. We found the preprocessing methods directly impact the concordance of the results. Thus, we devised a variance-stabilizing transformation (VST) by taking the advantage of the larger number of technical replicates available on the Illumina microarray. The VST can be extended to Affymetrix by utilizing the pixel variance of each probe. The MAQC-I data set was used to evaluate the results. Comparing to other popular methods, the VST transformation improves the cross-site concordance of both Illumina and Affymetrix microarray data. It also improves the cross-platform concordance between Illumina and Affymetrix. The algorithms are included in the lumi package of Bioconductor (www.bioconductor.org).
12:00 pm Pitfalls in Microarray Analysis that Contribute to the Need for Validation
Steven Enkemann, Ph.D, Director, Microarray Core Laboratory, H. Lee Moffitt Cancer Center and Research Institute
Validation means different things to different people and in different contexts. Is one validating a gene that was found on a microarray gene list? Or, perhaps validating the biological effect one observed in a microarray experiment? For a small experiment validation may be a mechanism for discriminating between noise and true differences. But, for a well-powered study it is now quite clear that microarrays produce reproducible and reliable results, so validation is not necessary to remove false positives. For most people validation means to verify the results from a microarray analysis. We will demonstrate several pitfalls in the concept of microarray analysis that lead to false conclusions and thus have contributed to the requirement for validation. These pitfalls include the fact that microarrays detect transcripts and not genes; the fact that an increase in probe intensity does not necessarily mean that a transcript has increased; and, the fact that although two probes are better than one, not all probes provide equivalent information.
12:30 Lunch On Your Own or Luncheon Technology Workshop
1:30 Chairperson’s Remarks
1:35 Multi-Lab Evaluation of Expression Profiling Platforms
Stephen Tirrell, Ph.D, Director, Molecular Technologies, Millennium Pharmaceuticals
Selecting laboratories capable of meeting specific requirements for data fidelity, quality and capacity requires rigorous assessment. We have conducted a multi-lab evaluation to assess their performance versus acceptance criteria in an effort to identify contract research labs capable of processing a large number of preclinical samples for biomarker discovery. The goals, experimental design and test results for microarray and RT-PCR platforms will be discussed.
2:05 Acceptance Criteria for Quantitation of Cytokine Gene Expression in Peripheral Blood Mononuclear Cells (PBMCs) by Real-Time Quantitative PCR
Angela Keightley, Ph.D., Principal Scientist, Molecular Biology, Laboratory Sciences, Charles River Laboratories
2:35 Regulatory Perspective
Bharat Joshi, Ph.D., Office of Cellular, Tissue and Gene Therapies (OCTGT), Division of Cellular and Gene Therapies (DCGT), Tumor Vaccines Biotechnology Branch, US Food and Drug Administration (FDA)
3:05 Networking Refreshment Break, Last Chance for Poster and Exhibit Viewing
3:30 From GeneChip to Taqman Low Density Array Comparing Microarray and PCR-based Gene Expression Measurements
Fred Immermann, M.S., Associate Director, Translational Medicine Biostatistics, Wyeth Research
Based on comparisons of Affymetrix GeneChip RNA expression levels measured in peripheral blood samples collected from large groups of diseased (n=337) and healthy subjects (n=348), a set of approximately 200 transcripts were identified for further evaluation by Taqman Low Density Arrays (TLDAs). The transcripts included products from genes that appear associated with disease, disease severity, and therapeutically relevant signaling pathways, as well as transcripts selected as controls for gender, cell composition, and normalization. The transcripts probed by the Affymetrix platform were mapped to one or more inventoried (standard, validated) assays available from Applied Biosystems, which were then arrayed onto two 96-well TLDAs. As one of the first steps in validating the performance of the TLDAs, 25 pairs of samples from the original Affymetrix profiling studies were run on the TLDAs, to examine the concordance between GeneChip and TLDA-based measurements of gene expression. Each sample pair was selected because it provided relatively large fold-differences in GeneChip expression between diseased and healthy subjects for many of the transcripts of interest. Approaches to normalization and the concordance between the GeneChip and TLDA fold-differences in expression will be described.
4:00 Leveraging Multiple RNA Measurement Tools to Support Discovery Research Programs
Shelley Ann Des Etages, Senior Principal Scientist, Genetic Technologies, Pfizer, Inc.
4:30 Using Hepatocytes to Underly the Mechanism of Species-Specific Toxicity of a PPAR Agonist
Yin Guo, Ph.D., Research Scientist, Exploratory Toxicology, Philip Morris USA
Using rat and dog primary hepatocytes, we replicated the pharmacologic and toxicologic effect of a novel PPAR agonist observed in vivo during preclinical development and propose an underlying mechanism for these species-specific effects based on gene expression analysis. This study utilized microarraysand quantitative PCR.
5:00 Panel Discussion with Afternoon Speakers
5:30 Close of Day