The simulation experiment description markup language (SED-ML): language specification for level 1 version 4

Abstract

Computational simulation experiments increasingly inform modern biological research, and bring with them the need to provide ways to annotate, archive, share and reproduce the experiments performed. These simulations increasingly require extensive collaboration among modelers, experimentalists, and engineers. The Minimum Information About a Simulation Experiment (MIASE) guidelines outline the information needed to share simulation experiments. SED-ML is a computer-readable format for the information outlined by MIASE, created as a community project and supported by many investigators and software tools. The first versions of SED-ML focused on deterministic and stochastic simulations of models. Level 1 Version 4 of SED-ML substantially expands these capabilities to cover additional types of models, model languages, parameter estimations, simulations and analyses of models, and analyses and visualizations of simulation results. To facilitate consistent practices across the community, Level 1 Version 4 also more clearly describes the use of SED-ML constructs, and includes numerous concrete validation rules. SED-ML is supported by a growing ecosystem of investigators, model languages, and software tools, including eight languages for constraint-based, kinetic, qualitative, rule-based, and spatial models, over 20 simulation tools, visual editors, model repositories, and validators. Additional information about SED-ML is available at https://sed-ml.org/.

Keywords

• computational modeling
• reproducibility
• simulation experiment