Overview
Newsworthy Events
- Tusher V, Tibshirani R & Chu G: Significance analysis of microarrays applied to the ionizing radiation response. Proc Natl Acad Sci USA 98: 5116-5121, 2001
- Tibshirani R, Hastie T, Narasimhan B, Chu G: Multi-class diagnosis of cancers using shrunken centroids of gene expression. . Proc Natl Acad Sci USA 99: 6567-6572, 2002
- "Peak probability contrasts for mass spec"
Proteomics data presents new problems of size and complexity that do not occur in genomics data from DNA microarrays. There are many more proteins than genes, the proteins can interact with each other in complex pathways, and each protein can undergo many post-translational modifications. Proteomics data includes separate data sets collected by a variety of different technologies. To extract information most effectively, these data sets must be merged with each other and with gene expression data, and then analyzed globally for biological effects and correlations.
Several methods have been developed specifically for the purpose of analyzing DNA microarrays for gene expression. These include unsupervised clustering algorithms and supervised search methods. As in the case of gene expression data, proteomics data involves very large data sets. However, unlike gene expression data, proteomics data must be generated with a number of different technologies. For example, protein levels, post-translational modifications, and protein interactions must be analyzed by different methods. Furthermore, the power of proteomics data will not be fully realized unless it can be combined with global profiles of gene expression. Thus, the full complement of proteomics data will include different units of measurement, ranges of expression, and levels of uncertainty. The technical challenge will be to devise methods for analyzing proteomics data, taking into account the size and complexity of the data sets.
Our goals for project 4 are to:
Develop methods to standardize, merge, and analyze proteomics data from
different data sets. We will develop methods to standardize and merge
proteomics data into a single megaset and then analyze the data for biological
effects and correlations.
Develop methods to diagnose murine and human autoimmune disease from
blood-derived proteomics data. Datasets from animal models and humans
with systemic lupus erythematosus (SLE) and rheumatoid arthritis (RA)
will be used to develop methods for using megasets of proteomics and
genomics data to classify samples to one of several disease states.
Develop and apply methods to utilize megasets of proteomic, genomic,
and clinical data for prognostication and monitoring response to therapy.
Megasets of proteomic and genomic data will be analyzed to identify profiles
with utility for prognostication, guiding therapy, and monitoring response
to therapy.
