American Chemical Society
Browse

BMSS2: A Unified Database-Driven Modeling Tool for Systematic Biomodel Selection

Download (4.46 MB)
journal contribution
posted on 2022-07-22, 12:08 authored by Russell Jie Kai Ngo, Jing Wui Yeoh, Gerald Horng Wei Fan, Wilbert Keat Siang Loh, Chueh Loo Poh
Modeling in synthetic biology constitutes a powerful means in our continuous search for improved performance with a rational Design–Build–Test–Learn approach. Particularly, kinetic models unravel system dynamics and enable system analysis for guiding experimental design. However, a systematic yet modular pipeline that allows one to identify the appropriate model and guide the experimental designs while tracing the entire model development and analysis is still lacking. Here, we develop BMSS2, a unified tool that streamlines and automates model selection by combining information criterion ranking with upstream and parallel analysis algorithms. These include Bayesian parameter inference, a priori and a posteriori identifiability analysis, and global sensitivity analysis. In addition, the database-driven design supports interactive model storage/retrieval to encourage reusability and facilitate automated model selection. This allows ease of model manipulation and deposition for the selection and analysis, thus enabling better utilization of models in guiding experimental design.

History