roadmap.md

ROAD MAP

Created: 2022-03-02

Updated: 2022-03-19

Methods

I. Types revision

• observables revision: compatibility with DiffEq, interpolation techniques instead of saveat approach
• Model revision: import from macros, general-purpose ODE consumption, faster loading
• mc (Monte Carlo) method updates: online-statistics, auto-stop by criterion

II. Optimization

• Parameter updates in Measurement
• [+] loss method revision: separate from fit method
• fit method revision: interface update, another optimization methods, visualization
• identify method: calculation of CI, CB

III. Visualization

• Extend plot recipes: compose plots based on tags, split plots horizontally

IV. Non-ODE simulations

• NonODEScenario implementation: allow simulation and fitting by explicit expressions like in DBSolve

V. Table formats

• extend Measurement format: more flexible tables, wide-format, macros
• PETAB compatibility: PETAB standardization, PETAB import

VI. Additional methods

• Parallel fitting methods: genetic, semi-parallel approaches
• Implementation of Global sensitivity analysis (GSA): FAST, eFAST, Sobol
• Graphical User Interface: Qt, VSCode plugin, web-application
• Markov Chain Monte Carlo (MCMC)
• Non-Linear Mixed Effect (NLME)

Techniques and use cases

• Analysis of dependent fitted parameter
• Validation
• Local sensitivity
• Global sensitivity
• Structural identifiability
• Practical identifiability and predictability
• Optimal experiment development (design)
• Predictive and prognostic biomarker discovery
• Virtual population
• Cloud resources for parallel simulations
• Working examples: dynamic reports, auto-reporting, web applications