A better understanding of the complexity of the role genetic and environmental factors play in the development of juvenile idiopathic arthritis (JIA) would greatly improve diagnosis and treatment of the disease. This would be particularly valuable if it led to the ability to predict which JIA patients could face severe joint erosion. At present, clinicians cannot predict this outcome in the early stages of JIA onset.

Studies in progress include a contract to identify the genetic components of JIA and a P01 grant study to identify gene expression patterns to use as novel disease biomarkers. These offer the potential for large-scale datasets of genetic, gene expression and clinical information, which will allow new and powerful insights into the biological pathways and processes that underlie JIA.

Our general aim is to develop novel computational methods for large-scale data mining and classification, and to build carefully validated databases of JIA gene expression and relevant polymorphism data. These resources would facilitate multidimensional mining with the ability to leverage the use of a priori knowledge of relevant gene-gene interactions and pathways.

We are in the process of building an integrated database of clinical, gene expression and polymorphism profiles for a large population of JIA patients. In conjunction with Jarek Meller, PhD, and Michael Wagner, PhD, mathematicians from the Division of Biomedical Informatics, we are constructing large-scale classification algorithms suitable for the analysis of diverse genomic and clinical data. We are placing special emphasis on dimension reduction and the development of statistical significance measures, two key computational challenges arising from these types of data. Once the database and computational tools are complete, we intend to use them to classify JIA subtypes and clinical outcomes.

Measuring the level of mRNA of all genes in blood cells may provide a molecular basis for treatment decisions at early stages of disease. We are using microarray analysis to measure this expression and identifying patterns specific to JIA patients, JIA subtypes, responses to treatment and outcome variables.

Additionally, the integration of data from gene expression and genetic analyses will help increase our understanding of disease. We intend to develop novel computational methods for large-scale data mining and classification. These methods will integrate both the gene expression and genetic data to facilitate multidimensional modeling of the biological pathways and processes that underlie JIA. In conjunction with clinical data, this analysis will yield powerful insights into complex pediatric diseases like JIA.

The Pediatric Rheumatology Tissue Repository (PRTR), a funded component of a P30 rheumatic disease core center, assists local and national investigators in the collection, processing, storage and distribution of biospecimens for rheumatic disease-related research.

The PRTR collects, processes and maintains specimens from patients with pediatric rheumatological and related musculoskeletal conditions and from relevant control populations. Samples include blood, urine, synovial fluid, synovial tissue, hair and saliva. Products derived from these collections are suitable for genetic, genomic, proteomic, immunologic function and biomarker studies.

Our collection, storage and processing practices are optimized to meet current and future study needs. Where possible, practices are standardized to match other national and international repository efforts. Likewise, informatics solutions and IRB guidelines are applied to sample storage information, allowing for accurate and appropriate use of samples. Our systems are optimized to allow researchers to consider consent, source of collection (study, external site, type of sample), and specifics of handling / processing as well as prior use when accessing samples.