The Das Lab

Research in Dr. Das’s laboratory focuses on the advancement of model-driven architectures and Semantic Web technologies (XML, OWL, SWRL) to support temporal reasoning, data integration, and collaborative systems in healthcare and the life sciences.

Research projects include:

1. Temporal knowledge discovery.

As the number and size of research data sets grow rapidly in biomedicine, many investigators urgently need computational methods that can enable them to manage complex results and to learn from those data. In research involving biological pathways, disease progression, and clinical trials, the ability to examine temporal relationships among study data can be critical to scientific inquiry. In collaboration with drug resistance, health services, and clinical immunology researchers, we are developing methods for ontology-driven database querying, statistical aggregation and data visualization.

2. Biomedical genomics research.

The clinical value of high-throughput technologies (such as gene microarrays) depends on adequately capturing the clinical context within which gene and protein expression data are collected. In collaboration with the Immune Tolerance Network—a NIH funded scientific collaboration among researchers studying immune tolerance mechanisms through clinical studies—we are creating computational methods to undertake large-scale, automated integration of mechanistic and clinical data. Our software approach to trials management and results analysis relies upon the use of ontologies to model patient information, clinical protocols, and mechanistic studies within a knowledge-based architecture.

3. Patient decision making.

Recent research indicates that treatments administered by physicians are frequently inconsistent with formally assessed preferences of individual patients. Health decision aids (HDAs) can help to make patients more involved in their healthcare choices, but their use in routine clinical practice faces a number of barriers, including the lack of time and resources available during an office visit. The dissemination of electronic medical records (EMRs) offers an opportunity to make HDAs more widely available. Little research currently exists on how to integrate HDAs into EMRs and how data on health preferences can be collected and stored for use in clinical care. In collaboration with Drs. Alan Garber and Mary Goldstein, we are developing a knowledge-based method, called Health e Decision, to enable decision models to be driven by data in EMRs and to be accessible through patient web portals.