research projects
Many but not all of the affiliated research projects are listed below. See also "Featured Research"
Bio-STORM
Principal Investigator: Mark Musen
The Bio-STORM project involves surveillance of publicly available databases
to detect patterns suggestive of incipient epidemics, possibly the result
of potential acts of bioterrorism. Research addresses problems of data
integration, high-performance problem solving, and elucidation of new
analytic approaches for pattern detection.
National Center for Biomedical Ontology: cBio
Principal Investigator: Mark Musen
How do humans reason? We create mental models of the problem using concepts that are expressed mathematically as symbols and figures or as spoken and written languages such as English. Before we can expect to automate any human problem-solving process, we must express the problem in terms that the computer can manipulate and which the human community can understand. We create a model of our body of expertise (biomedical concepts), known as the knowledge representation and duplicate our reasoning processes on this body of knowledge using algorithms (simulation, probabilistic reasoning, artificial intelligence, machine learning, etc...) in order to produce software that automates one of our problem-solving procedures. This process lies at the heart of every reasoning task which has been automated.
An ontology or knowledge representation is a formal "specification of a conceptualization" . As a simple example of an ontology, we may visualize a cell as a collection of macromolecules (proteins, lipids, nucleic acids, etc...). Therefore, providing the name, location in space and chemical identity of all macromolecules should be sufficient to specify our simple conceptualization of a cell. The development of an appropriate ontology/knowledge representation has historically been an integral part of the software development process. As our body of knowledge grows in depth and complexity, it becomes difficult for one team of software engineers/biomedical domain experts to maintain the knowledge representation to keep current software implementations running. The National Center for Biomedical Ontology (cBIO) mission is to develop re-useable, communal ontologies whose breadth and scope span the field. See also "Featured Research" for more on the cBio project.
Collagen: Linking Collagen Genotypes to Molecular Phenotypes
Principal Investigator: Teri Klein
Collagen is the most abundant
protein in humans, and mutations in collagen can lead to death or disease. The
ways in which individual mutations affect collagen structure/function are poorly
understood. Collagen is a fibrillar protein, and thus many of the principles
elucidated for the study of globular proteins are not immediately applicable
in investigating the relationship between its structure and function. We now
have an opportunity to use genomic technologies to survey the variation in key
collagen genes throughout the human population, link the discovered polymorphisms
to their structural effects, and develop an understanding of the mechanism of
collagenous disorders.
Chronus
Primary Contacts: Amar Das, Martin O'Connor, Ravi Shankar
The Chronus project is a long-term research program at Stanford Medical
Informatics to advance temporal database theory and temporal querying
methods for software applications in clinical care and biomedical research.
The open-source Chronus system is currently being evaluated in the context
of guideline-based decision support and HIV drug resistance research.
E-Preference
Principal Investigator: Amar Das
The goal of the e-Preference project, which is funded by Stanford's Center
on Advancing Decision Making for Aging, is to use ontologies and influence
diagrams as the software foundation for patient-centered decision aids.
The e-Preference tool integrates knowledge from domain experts with patient
preferences to support shared decision making in clinical care.
EON
Principal Investigator: Mark Musen
The focus of the EON project is the automation of best practices in medicine
and evaluation of knowledge-based techniques to enhance clinician adherence
to clinical practice guidelines and clinical protocols.
FEATURE
Principal Investigator: Russ Altman
FEATURE is a research program for characterizing and recognizing active
sites, binding sites, and other sites of functional/structural importance
in macromolecules. It's goal is to annotate new protein structures with
functional information, in order to assist in creating a link between
biological structure and function.
Konark
Principal Investigator: Amar Das
The Konark project is focused on the development of an integrated framework
to combine database querying, data abstraction, and data and text mining
methods for temporal knowledge discovery and integration in biomedical
research domains, particularly those involving analysis of time-course
data in translational bioinformatics.
PharmGKB
Principal Investigator: Russ Altman
The PharmGKB is the Pharmacogenetics & Pharmacogenomics Knowledgebase.
Pharmacogenomics is the study of how variation in the human genome contributes
to variation in drug response. The PharmGKB stores information about genes,
drugs, diseases, drug-related pathways, drug-related phenotypes, and the
literature that describes current knowledge.
Protege
Principal Investigator: Mark Musen
Protégé is the most widely used system in the world for creating and managing
online biomedical knowledge bases. Current work involves development of
enhanced methods for using Protege to construct knowledge-based systems,
including reusing problem-solving components, simplifying conceptual modeling,
and developing new applications for the Semantic Web.
Simbios
Principal investigator: Russ Altman
Simbios is a National Center for Biomedical Computation focused on physics-based simulation of biological structure--from molecules to entire organisms. It has two goals: to perform innovative research in methods for simulation and analysis of biological dynamics, and to disseminate tools and models to the entire biomedical research community, through the SimTK site. See also the "Featured Research" page.