John Gennari
University of Washington, WA·Computer Science
Computer models of biological systems, called biosimulation models, are accumulating in public repositories. They are only valuable if other scientists can find, understand, rerun and trust them, and in practice reproducing a published model or judging its credibility is often difficult.
- Fields
Computer Science
Biomedical Engineering
- Topics
biosimulation
computational modeling
reproducibility
- Grants
- 1 known grant, through Mar 2029
What they work on
- Develop standards and ontologies, structured vocabularies of defined concepts, that capture the semantics of biosimulation model evolution and versioning so users can select the appropriate model version.
- Create standards and ontologies for model composition and decomposition that work even when the component models were developed with different mathematical paradigms.
- Build tool support that automatically recommends semantic annotations for models, lowering the effort required to describe what model components mean.
- Extend standards and repositories that capture model execution, including new development of the SED-ML standard for describing simulation experiments.
More about the research
The community largely knows what makes models reproducible; what it lacks are the repositories, standards and tools to put those ideas into routine practice.
Why it matters
Scientists who build or reuse computational models of physiology and other biological systems could gain practical ways to find, interpret and rerun existing work. Shared annotation standards, automated annotation tools and records of how simulations were run could make biosimulation results more reproducible and credible, so models are built upon rather than rebuilt.
Known grants
From NIH and NSF research grants. Labs often have other funding we can't see.
Closest labs
Piper Ranallo
University of Minnesota, MN·Computer Science·Early-career
Build machine-readable ontology models that define mental, behavioral and social processes, findings, assessment instruments and interventions, using borderline personality disorder and psychosis as test cases.
Lavanya Elluri
Texas A&M University-Central Texas, TX·Computer Science
Develop a technique to automatically extract rules from U.S. IoT cybersecurity laws, data protection regulations and NIST standards, and store them in semantically rich ontologies built with Semantic Web technologies.
Junfeng Yang
Columbia University, NY·Computer Science
Detect model spoofing, where attackers upload malicious AI models under deceptive names, by developing anomaly detection based on naming conventions and model architectural signatures.
The school
University of Washington: 81 labs on PI Pages, $1.7B a year on research, 694 research doctorates a year.
Reach out
A start for your email. Fill in the brackets.