Tutorial details

Date: Saturday September 3, 2016
Time: 9:00 – 12:30
Venue: World Forum, room: Antarctica

Tutors

• Andrea S. Foulkes, Mount Holyoke College, Massachusetts, United States

Summary

Advances in high-throughput technologies and the pursuant accumulation of a vast array of new sequence and -omics level data offer new opportunity to gain insight into complex disease etiologies. Moreover, for complex traits, we increasingly recognize that the intergenic space between protein-coding genes (PCGs) – previously thought to be non-instrumental – contains highly ordered regulatory elements that control expression and function of PCGs and themselves can be actively transcribed molecules. Analytic paradigms that integrate and combine information from these emerging taxonomies and data elements are essential for focused inferential interrogation in this setting. This tutorial focuses on specific methods and R tools designed to uncover these complex, and largely uncharacterized relationships.

The overall goal of this tutorial is provided participants with fundamental knowledge, resources and a framework for staying abreast with novel developments in the field of statistical genomics. Upon completion of this tutorial, students will: (a) understand the core analytic components of a standard genome-wide association (GWA) study and be familiar with code to complete a “start-to-finish” GWA analysis using R; and (b) understand and be able to apply approaches for more comprehensive interrogation of regional and higher-order associations to discern the relative contributions of gene and regulatory elements.

Target audience

This tutorial is intended for students and practitioners interested in learning fundamental core concepts and R tools for high-dimensional genomic investigations and post-analytic interrogation. Participants are expected to be familiar with basic statistical concepts and techniques at a level of an introductory course in statistics.

Detailed Schedule

09:00 – 10:30 Part 1: Core analytic components of a standard GWAS

• Data preprocessing – SNP and sample level filtering;
• Data generation – principal components for population substructure and SNP imputation using 1000 genomes data; and
• Association analysis – for typed and imputed SNPs using parallel processing

 
10:30 – 11:00 Coffee Break

11:00 – 12:30 Part 2: Interrogation of regional and higher-order associations

• Post-analytic visualization and interrogation using external resources, including those available through the UCSC genome browser;
• General overview of class-level testing strategies, rare-variant analysis, gene-environment interaction analysis, and gene-set enrichment and biological pathway analysis. Specific methods covered will include: SKAT, GATES, VEGAS, QT and GenCAT;
• Error control, computational efficiencies and versatility of each method with respect to alternative genomic taxonomies, including regulatory elements such as long intergenic non-coding RNAs, enhancer elements and splicing regions.

Requirements & material

Participants are encouraged to bring their laptops. All R/jupyter slides will be made available, in addition to detailed instructions for required R packages and data downloads, at the following URL by mid-August: http://www.stat-gen.org/.