Statistical methods for High-throughput Genetic Data (07w5023)

Objectives

The overall objective of this workshop is to bring together statisticians, geneticists, molecular biologists, and scientists in other related disciplines to exchange ideas and problems with the ultimate goal of establishing productive collaborations and developing efficient statistical methods and study designs to address scientific questions arising from genomics studies. Because of the very broad scope of genomic studies, we will focus on the analysis of DNA variations (SNPs mostly well-known in this class) and their impacts on biological systems at different levels (such as gene expressions and disease phenotypes) to ensure that the workshop is focused enough to be productive yet diverse enough so that most ongoing genomics topics are included.
The participants will come primarily from two backgrounds: quantitative scientists interested in utilizing their skills to solve critical biological problems and biological researchers in search of efficient computational and statistical methods to help with their research.

We will maintain a well-balanced mix between well-seasoned researchers in the covered fields and those wanting to learn and contribute to these areas.

Relevance: Although this workshop will be mostly organized around scientific problems, namely, DNA variations and their effects on biological systems at various levels, the central theme that is cutting through all topics is the developments and applications of efficient statistical and computational methods to address these questions. Due to the complexity of biological systems and the amounts/types of available data, existing statistical methods are not readily applicable to these problems. Therefore, novel statistical methods under existing frameworks, e.g. likelihood-based inference or Bayesian methods, may be developed. However, it is even more likely that new statistical framework/formulation needs to be putted forward under which context to consider and solve these problems. These efforts and goals are only possible with close interactions among leading scientists from quantitative and biological fields, and the proposed workshop will provide a great platform for such interactions and joint work.

Importance: The scientific importance of the topics to be covered is very apparent. With the accumulation of large data sets of distinct types, the contributions of statisticians are more fully appreciated by biologists and medical resarchers. However, it is impossible for statisticians to work in a vacuum or in isolation to develop useful statistical methods. One such example is the analysis of microarray data. Although much has been published in the statistical literature, most work has focused on minor statistical issues and will likely have little impact on real biology. Therefore, this workshop is important in that it provides a forum for statisticians to mix with biologists allowing both groups to identify important scientific questions.

Timeliness: A workshop centered around DNA variations and their effects is very timely as most human genetic variation data will be gathered by 2007 and it is immensely important to understand how these variations affect us at different levels, how evolutions have shaped these variations over time, and how they reveal the underlying biological systems.