Smart searching for the Physiome Project

The International Union of Physiological Sciences (IUPS) launched the Physiome Project at the IUPS World Congress in St Petersburg in 1997 in order to bring quantitative bioengineering approaches to the study of anatomy and physiology. In particular this approach advocates the use of anatomically and biophysically based mathematical models that deal with physiological processes operating over multiple spatial and temporal scales. The major focus of the project has been the creation of standards, tools and databases for reproducible and reusable mathematical models of physiological processes (www.physiomeproject.org).

We now have well established standards for encoding mathematical models and simulation experiments, with the Physiome Model Repository (PMR, models.physiomeproject.org) containing over 600 models. The focus has recently shifted to developing technologies for being able to semantically describe the models and simulations in order to help scientists understand the actual biology represented by the model, suitable applications for the models or simulations, provenance of the knowledge contained in the repository, etc. Using these technologies, we now have a growing collection of knowledge contained in the PMR unambiguously defined in a computable manner based on standard semantic web technologies.

The goal of this project is to develop novel tools to enable discovery of relevant knowledge contained in this existing knowledgebase. We imagine this will involve using various language processing and text-mining technologies to translate user queries into specific semantic queries to be iteratively executed against the knowledgebase to discover the computational models and simulation experiments that most closely match the user requirements. The ultimate goal is a “Google search” style interface where the user can enter a simple text query in order to find models and simulations which are relevant to their work.

Key skills: students wanting to work on this project should have a background in biomedical engineering, computational/information sciences, semantic web, or similar; some programming experience would be useful.