Identifiability problems in systems biology
August 19 to August 23, 2019
at the
American Institute of Mathematics,
San Jose, California
organized by
Marisa Eisenberg and Nicolette Meshkat
Original Announcement
This workshop will be devoted to identifiability problems in systems biology. Identifiability is the problem of determining which unknown parameters of a mathematical model can be determined from known input-output data (i.e. from particular variables which are measured/observed). In the structural identifiability problem, data is assumed to be perfect, i.e. noise-free and of any time duration required. Structural identifiability is a necessary condition for the numerical or "practical" identifiability problem, which is the parameter estimation problem for real and often noisy data. A lack of appreciation of parameter identifiability and uncertainty has been pointed to on numerous occasions as hindering the progress of mathematical modeling in biology.
We hope to address identifiability problems in both the theoretical and application-driven side of mathematical modelling. We anticipate that this work will lead to new collaborations, formulation of scientific questions as precise mathematical problems, and the introduction and sharing of new algebraic and analytic techniques to address inference challenges in mathematical biology. Among the problems we may consider are the following:
- Algorithms and Computation
- Symbolic computation issues and the global identifiability problem
-
Efficient algorithms to test local identifiability
- Identifiable Combinations and Geometric Structure
- Parameter space reduction and identifiable reparametrizations of unidentifiable models
- Computing identifiable combinations of parameters in unidentifiable models
- Applications in Practice
- The numerical identifiability problem and parameter estimation
- Examining the consequences of unidentifiability on control strategies and predictions made from models in practice
Finally, we will synthesize the above problems to identify several key conjectures, problems, and open questions in the area of identifiability analysis of dynamic systems models. These problems will be both mathematical questions, as well as application-focused questions.
Material from the workshop
A list of participants.
The workshop schedule.
A report on the workshop activities.
Workshop Videos
Papers arising from the workshop:
Identifiability of linear compartmental tree models
by Cashous Bortner, Elizabeth Gross, Nicolette Meshkat, Anne Shiu, and Seth Sullivant
Model structures and structural identifiability: What? Why? How?
by Jason Whyte
Branching out into Structural Identifiability Analysis with Maple: Interactive Exploration of Uncontrolled Linear Time-Invariant Structures
by Jason Whyte
Computing all identifiable functions for ODE models
by Alexey Ovchinnikov, Anand Pillay, Gleb Pogudin, Thomas Scanlon
Multi-experiment parameter identifiability of ODEs and model theory
by Alexey Ovchinnikov, Anand Pillay, Gleb Pogudin, Thomas Scanlon
