The Carolina Center for Exploratory Genetic Analysis (CCEGA) is developing an interdisciplinary infrastructure to identify the complex genetic traits that underly human diseases, bringing together data from clinical studies, population studies and model systems. CCEGA believes the next breakthroughs in our understanding of biology and disease will be made possible by the integrated analysis of genetic data and its expression as phenotypes. CCEGA work centers on enabling this kind of multidisciplinary, multi-investigator research. The center involves three complementary groups of scientist at the University of North Carolina at Chapel Hill: (a) experimental geneticists, (b) quantitative experts in statistics and biostatistics, and (c) computer scientists with expertise in algorithm development, software construction, and high-performance computing.

Phase one of CCEGA focuses on building a community of investigators and deploying a  prototype infrastructure for analyzing relationships among genotypes and phenotypes in three contexts:

• Family linkage studies, which examine the relationship between genotypes and  susceptibility to specific diseases and conditions, in this case alcoholic addiction.
• Gene expression profile studies, which develop a picture of genes and cellular activity in order to identify patterns and signatures related to disease, in this case breast cancer.
• Public health studies, which look at communities and their risk factors for diseases, in this case atherosclerosis.

The RENCI Contribution
To accommodate the diverse, multi-investigator databases necessary to answer these complex questions, RENCI is working with scientists to develop a prototype, extensible data model and provide access to data via a portal constructed using the Open Grid Computing Environment toolkit. The newest methods of integrated data analysis will be incorporated into a portal-based workflow. These include new techniques in linkage analysis (oligogenic analysis, multivariate linkage analysis, epistasis, and genotype by environment interaction), subspace clustering, and association analysis (quantitative trait and nucleotide analysis).

RENCI and its scientific partners also are exploring new visualization techniques for examining and interacting with large data sets and high performance computing for implementing computationally intensive analysis techniques. To reduce the barriers between data providers and data analyzers, CCEGA and RECNI conducts intensive, specialized workshops, colloquia and intramural meetings.

Funding
National Institutes of Health/National Center for Research Resources, Grant Number 5-P20-RR020751-01-02

Co-Principal Investigators at UNC-Chapel Hill

James Evans, Terry Magnuson, Karen Mohlke, Fernando Manuel Pardo, Charles Perou, Patrick Sullivan, David Threadgill, Kirk Wilhelmsen, Department of Genetics
Susan Paulsen, Jan Prins, Wei Wang, Department of Computer Science
Fred Wright, Fei Zou, Department of Biostatistics
Bradley Hemminger, School of Information and Library Science
Andrew Nobel, Department of Statistics
Kari North, Department of Epidemiology
Alexander Tropsha, School of Pharmacy
K.T.L. Vaughan, Health Sciences Library

RENCI Team

Xiaojun Guan
Kevin Gamiel
Clark Jeffries
Jeff Tilson

Publications

Fred A. Wright, Hanwen Huang, Xiaojun Guan, Kevin Gamiel, Clark Jeffries, William T. Barry, Fernando Pardo-Manuel, Patrick F. Sullivan, Kirk C. Wilhelmsen, and Fei Zou. Simulating Association Studies: a Data-based Resampling Method for Candidate Regions or Whole Genome Scans (accepted for publication in Bioinformatics), 2007.

Jeffries, C. Hairpin Database: Why and How? Genomic Impact of Eukaryotic Transposable Elements  conference, Asilomar, CA, April 2006

Jeffries, C. Bipartite and tripartite systems and matrices from genetic control research, Linear Algebra and its Applications 409 (2005) 70-78.

Jeffries, C., Jarstfer, M., Perkins, D.: Folded RNA from an intron of one gene might inhibit expression of a competing gene, in silico Biology 5 (2005), 0037.

Jeffries, C., Perkins, D., Jarstfer, M.: Systematic discovery of the grammar of translational inhibition by RNA hairpins, Journal of Theoretical Biology (accepted for publication).

J. Liu, S. Paulsen, X. Sun, W. Wang, A. Nobel, J. Prins, "Mining Approximate Frequent Itemsets In the Presence of Noise: Algorithm and Analysis", SIAM Conference on Data Mining (SDM), 2006.

J. Liu, S. Paulsen, W. Wang, A. Nobel, J. Prins, "Mining approximate frequent itemset from noisy data", Proceedings of the 5th IEEE International Conference on Data Mining (ICDM), 2005.

Hemminger BM, Saelim B, Sullivan PF. TAMAL: An integrated approach to choosing SNPs for genetic studies of human complex traits. Bioinformatics 2006.

Presentations

 

Partners
RENCI
University of North Carolina at Chapel Hill :

Department of Biostatistics
Department of Computer Science
Department of Epidemiology
Department of Genetics
Department of Statistics
Health Sciences Library
School of Information and Library Science
School of Pharmacy

Activities

Working Groups
There are three working groups that meet every week to have discussions and presentations on specific topics.

• Informatics Working Group - Meets 12:30 - 1:30 every Thursday, contact: Bradley Hemminger, School of Information and Library Science, bmh@ils.unc.edu, 919-966-2998.
  
  

  Workgroup Presentations

  PCaP-Epidemiology Specimen Tracking System	Roger Akers, Feb. 17, 2005
  Lab Data Management Systems	Kirk Wilhelmsen, Feb. 24, 2005
  Demo of Lab Data and Clinical Data Management Systems	Kirk Wilhelmsen, Mar. 3, 2005
  Generalized Model (Modeling Genetics and Proteomics studies)	Brad Hemminger, Mar. 10, 2005
  Review Draft Model	Brad Hemminger, Mar. 24, 2005 (meeting minutes).
  Review Draft Model	Brad Hemminger, Mar. 31, 2005 (meeting minutes).
  Review Draft Model	All, Apr. 7, 2005 (meeting minutes).
  BSP (BioSpecimen Project) Facility	Peter DeSaix, Paul Brown, May 19, 2005 (meeting minutes).
  Knowles lab and their databases related to cystic fibrosis (CF)	Mike Knowles, Hemant Kelkar, Annie Xu and David Fargo, May 26, 2005 (meeting minutes).
  Identify genes that influence cardio vascular disease	Kari North, Jun 2, 2005 (meeting minutes).
  Data management issues of Melanoma project	Dennis Simpson, July 7, 2005 (meeting minutes).
  Compare and merge database schemas of UNC labs	Offsite meeting at the Friday Center, Nov 8, 2005
  Review initial draft of the schema for the common data model	March 24, 2006 (meeting minutes).

• Analysis Working Group - Meets 11:00 - 12:00 every Tuesday, contact: Jan Prins, Department of Computer Science, prins@cs.unc.edu, 919-962-1913.
  
  

  Workgroup Presentations

  Linkage Analysis	Ethan Lange, Feb. 3, 2005
  Family Based Association Studies	Kirk Wilhelmsen, Feb. 8, 2005 (summary).
  A Model of Genetic Data	Fred Wright, Feb. 22, 2005 (summary).
  A Model of Genetic Data	Fred Wright, Mar. 8, 2005 (summary).
  Survey of Data Mining Techniques for Genotype-Phenotype Association Studies	Pat Sullivan (slides) and Susan Paulsen (slides), Mar. 22, 2005 (summary).
  Survey of Data Mining Techniques for Genotype-Phenotype Association Studies	Pat Sullivan (discussion), Mar. 29, 2005
  Quantitative Genotype Phenotype Relationships (QGPR): Can we learn from Quantitative Structure Activity Relationships (QSAR) modeling?	Alex Tropsha (slides), Apr. 5, 2005 (summary).
  Application of neural networks to find fucntions of selected, weighted combinations of measurable parameters	Clark Jeffries (slides), Apr. 12, 2005 (summary).
  Microarray Data and Analysis	Charles Perou (slides), Apr. 19, 2005 (summary).
  Classification Accuracy Criteria as Target Functions in QSAR	Alexander Golbraikh (slides), Apr. 26, 2005




• ELSI Working Group - contact: James Evans, Department of Genetics, jpevans@med.unc.edu, 919-966-2276.

Tutorials As part of the Carolina Center for Exploratory Genetic Analysis (CCEGA), we have organized a tutorial series. The goal of the series is to provide cross-disciplinary education and facilitate interdisciplinary collaboration.

Tutorial Presentations (slides and audio)

Genotyping, (slides), (slides and audio)	Bob Millikan, Apr. 19, 2005
XML, (slides), (slides and audio)	Barrie Hayes, Apr. 19, 2005

---

Data Sets

We provide simulated data sets generated according to certain genetic models to allow computational scientists to evaluate various algorithms for mapping genotype-phenotype relationships. Developmental model 1: Diffusion-Threshold Two different sets of "Case-Control" data are available. In the Null data, there is no connection between phenotype and genotype. In the Developmental Model 1 data, phenotype is determined by genotype through a simulated developmental process dependent upon a subset of the genetic loci. The two types of data sets share the same population genetic parameters. Each collection includes 1000 files, one for each replicate data set. Each collection is packaged into three zip files for easy downloading, and one sample from each collection has been provided:

• Null data - description, an example, and the whole collection: null1.zip null2.zip null3.zip
• Developmental Model 1 - description, an example, and the whole collection: real1.zip real2.zip real3.zip

Contact: Susan Paulsen, Department of Computer Science, paulsen@cs.unc.edu

---

Participants

---

Acknowledgements

The project described was supported by Grant Number 5-P20-RR020751-01-02 from the National Center for Research Resources. Its contents are solely the responsibility of the authors and do not necessarily represent the official views of the National Institutes of Health/National Center for Research Resources.

Introduction
HAP-SAMPLE is a web application for simulating SNP genotypes for case-control and affected-child trio studies by resampling from Phase I/II HapMap SNP data. The user provides a list of SNPs to be "genotyped," along with a disease model file that describes causal SNPs and their effect sizes. The simulation tool is appropriate for candidate regions or whole-genome scans.

Acknowledgements
This project is supported by Grant 5-P20-RR020751-01-02 from the National Institutes of Health Center for Research Resources as part of the Carolina Center for Exploratory Genetic Analysis. Other sources of support include Carolina Environmental Research Center (EPA RD-83272001), NIGMS R01 GM074175, and CF Foundation Zou05P0. Its contents are solely the responsibility of the authors and do not necessarily represent the official views of the NIH or the National Center for Research Resources.

Link to www.hapsample.org