My research is a multidisciplinary computational approach to improve healthcare. I focus on ways to create technology that improves our understanding of healthcare and the life sciences in order to make medical practice more efficient. A more efficient process will make it easier for patients to receive safe and effective treatments at lower cost.
Health care costs have been rising for several years. The United Sates government reports that expenditures in the US on healthcare surpassed $2.3 trillion in 2008, more than three times the $714 billion spent in 1990, and over eight times the $253 billion spent in 1980. Reducing the cost of health care has become a national priority, as the government, employers, and consumers struggle to keep up with rising health care costs. My research aims to reduce human suffering and healthcare costs by focusing on treatment efficacy and safety.
Semantic Web Applications in Health Care and Life Science
The Semantic Web advances the World Wide Web with enabling technologies that make it possible for the meaning (semantics) of information on the web to be machine-readable. Over the past decade I have played a leading role in the definition and development of Semantic Web applications to the life sciences. As an organizer of the BioPathway Consortium (BPC), an established conference venue for emerging biopathways research, I obtained funding from the Department of Energy for BioPAX, an initiative to create a standard for biopathways data. BioPAX, an early adopter of OWL, the Web Ontology Language, has since become the international standard for biological pathway data, with the major pathway databases exporting in BioPAX format.
I explored semantic technology for use with drug discovery in collaboration with Dr. Eric Neumann, in which I modeled a cellular signaling process (Wnt pathway) using BioPAX (OWL). This was essential for the BioDASH prototype, which automatically combined chemical compound data, biological target (protein) data, and pathway data to reveal previously unknown targets for drug action. Sir Tim Berners- Lee, inventor of the World Wide Web and Director of the World Wide Web Consortium (W3C), used BioDASH in a BioIT World Keynote to demonstrate the utility of the Semantic Web for the pharmaceutical industry. Later, during his keynote at the International Semantic Web Conference, Sir Berners-Lee used my research to show the use of Resource Description Framework identifiers (RDF IDs) as database links to query over heterogeneous bioinformatics data sources that reside at different locations on the web. Beyond my computer-based research agenda, I made several presentations to the W3C to advocate for a W3C initiative in the Health Care and Life Sciences; the W3C HCLSIG is now in its fifth year.
Infectious Disease - Influenza Ontology
For the past four years I have been leading a collaboration with the University of Maryland (Gemina project) and
the University of Texas Southwestern Medical Center (BioHealthBase project) that extends the Infectious Disease Ontology to support influenza research, surveillance and outbreak monitoring. Last year the Canadian government joined the collaboration.
Ontologies, a formal representation comprising a vocabulary of terms and how they interrelate, have become popular among many disciplines because they make data sharing, integration and reuse much easier. However, the area of ontology evaluation, i.e. how one can objectively assess an ontology or part of an ontology for use in another application has been given little attention. Current methods are specific to a use case, rely on a qualitative evaluation by panel of experts who review the ontology, or are ranked by user votes within a community of practice. These methods, while present, provide little utility in practice and define no objective criteria for evaluation. To address this, I developed an approach that provides a framework within which existing methods and metrics are placed, and identifies the areas for which further research is needed. I used examples from existing ontologies to demonstrate how the evaluation process would work. The MITRE Corporation (MITRE), National Institute of Standards and Technology (NIST), National Center for Biomedical Ontology (NCBO), and several members of the ontology community, favorably reviewed this work. MITRE provided the funding for the initial research (2008) and NIST award 2012-2015 Award No. 60NANB12D20.
Systems Biology and Medicine – Major Depressive Disorder (MDD)
Although antidepressant treatment is readily available, the US government reports that Major Depressive Disorder is the leading cause of disability in the U.S. for ages 15-44. Why? Depression is complex. The diagnosis for Major Depressive Disorder describes the signs of depression, but not the cause. There are several distinct biological, physiological, and psychological systems that, when not functioning properly, could result in depression. Furthermore, clinicians are unable to predict which treatment will be successful, since not all patients respond to all antidepressant treatments. In order to properly treat depression, we need to understand the different types of depression, their underlying mechanisms of action, and how they interact in all three aspects: (1) behavioral (emotional and cognitive processes), (2) functional (chemical pathways from brain region to brain region), and (3) mechanistic (dynamics of the neurotransmitters and their receptors). To study the effects of different antidepressant treatments, how they affect the underlying brain regions, and how they manifest in clinical symptoms, I use mathematical modeling and computer simulation as a way to combine clinical research with neuroscience research. So far, my methods have been awarded two US Patents, have been licensed by commercial companies (they remain free for academic and non-profits), and have indicated, that different treatments result in different response patterns, i.e. with different treatments, the individual symptoms of depression (i.e. anxiety, mood, sleep disturbance, etc.) improve at different times and in a different order. Findings like these have the potential for huge clinical impact; they could help reduce patient suffering and suicide by reducing the trial and error of current practice.
Systems Biology and Medicine – Diabetes Type I and Type II
This research project has two goals, one narrow and one broad. The narrow goal is to explore the creation of a consistent model of insulin resistance in skeletal muscle. The model will be based on experimental data (microarray, mass spectroscopy) from the Joslin Diabetes Center and the University of Michigan, and will use a number of emerging technologies, including workflows (Taverna), visualization environments (CellDesigner), ontologies (OWL) and reasoners (Pellet). The research will investigate whether the representational technologies available are adequate to integrate the experimental data provided by the clinical researchers and whether the experimental techniques need refinement because the data are lacking in their information content. The outcome of this research, apart from the model, will be a deeper understanding of insulin resistance and an indication of where to focus further experimental and computational effort. The model will form the basis for continued work by enabling diabetes researchers to validate existing hypotheses, better predict experimental results, and generate new hypotheses to guide further experiments. The broad goal of this research is that the approach will have substantial applicability elsewhere in the life sciences.
My background is in mathematics and computer science with a PhD in Cognitive and Neural Systems. I am interested in modeling scientific and health/biomedical related knowledge in order to make it more useful, i.e. computationally reusable by other scientific research, or in practice, such as in analysis or decision support systems.
My PhD research focused on mathematical models of unipolar depression that would provide insight into the underlying dynamics of depression in order to support clinicians, enabling them to make better treatment choices, drug or therapy, for their individual patients. This type of approach is now known as personalized medicine. I used neural networks and differential equation modeling in this work. These approaches are known as machine learning and systems biology, respectively.
Translational Medicine - from Clinical Research to Clinical Practice
While researching in this area, I became aware of the need to integrate knowledge from and across multiple disciplines in order utilize the knowledge within each domain from its unique perspective. I studied the use of ontologies for knowledge representation and as a way to further exploit description logics (the formalism underlying the OWL ontology language) as a mechanism in which the integration of heterogeneous life science data could be automated. I focused this initial work in the area of biological pathways (signal transduction, metabolic, gene regulatory, protein interactions, etc.) which apply to a number of disease areas; for example, biological pathways form the basis for identifying and relating chemical compounds with putative targets for drug therapies. In support of this theme, I took on the role as community liaison for the BioPathways Consortium and co-led the development of a community-based standard for the exchange of pathway information (BioPAX) leading to the establishment of the BioPAX initiative, a global consortium within which the standards work takes place.
The seamless integration of clinical research, outcomes and life science knowledge, remains the major challenge in being able to reap the rewards of our nations' research investments in biomedical research.
Throughout my research career, I have sought the engagement of clinicians and scientists from multiple disciplines. In fact, identification of the so-called stakeholders in the first step. In order to provide any solution one must first understand the problems and the issues that domain experts face. This is not possible without engaging with others - as no one is an expert in all areas.
The Methods from my dissertation research, which were subsequently patented were sold in 2009 in order to be put into use in clinical practice.
In addition, I have a long-standing interest in teaching that I am looking to continue/develop further. Until now multidisciplinary research has meant that one finds oneself in departments where there isn't a direct fit and teaching isn't an option. Now with the growing recognition of the utility of in silico methods in enhancing clinical practice (the translational medicine movement), there is a better understanding of the contribution of each discipline in medical practice and much more support for it. Thus I look forward to being able to teach what I've learned in a multidisciplinary environment.
Biological Pathways Integration, Aggregation, and Inference
EMPWR: Computational Exploration of Molecules in the Context of Biological Pathway Networks \\ National Science Foundation Award Number IIS-0542041
(http://www.biopaxwiki.org). Core Group Member involved in intimating the creation of a data exchange format for biological pathway databases. Procured partial support from the DOE and got it doubled in the second year. Represents BIOPAX at international conferences and workshops and coordinate development with the BioPathways Consortium and its constituents. (Since August 2002)
Major Depressive Disorder (Unipolar Depression)
My graduate work was in the area of predictive medicine and personalized medicine. I was awarded two patents based on my graduate work, one for predicting the therapeutic outcome of a treatment and the other for an automated treatment selection method. This work was based on analyzing clinical research and clinical trial data for antidepressant treatments including cognitive behavioral therapy.
I have a strong interest in integrating diverse sources of knowledge, data (from multiple technologies) and various modeling and analysis methodology to address clinical, medical, biological and scientific questions. I welcome opportunities to collaborate in any disease area.
I have a strong interest in education and in collaborative work. This is evidenced by my emphasis on including tutorials as introductory parts of workshops and my involvement in creating consortia and collaborations in health care and life sciences.
(http://www.biopathways.org) Co-organizer and Steering Committee. The mission of the BioPathways Consortium is to catalyze the emergence and development of computational pathways biology, by building up a strong and coherent scientific community, sharing knowledge, facilitating collaborations, and fostering the development of methods and tools of wide interest to the community.
W3C Semantic Web for Health Care and Life Science (HCLSIG)
Dean's Advisory BoardBoston Univeristy Metropolitan College. Here's a photo from the meeting in November 2003 at Student Village Boston University and one from the Dean's Advisory Board Dinner, Veronique, Boston March 13, 2003