Table 1.

Criteria and considerations for adapting models.

Criteria	Considerations
Scope	• Does the model:
	° Represent appropriate biology?
	° Include necessary biological components and processes?
	° Include appropriate level of biological detail (especially for your target areas)?
	° Represent the appropriate timeframe (e.g., minutes vs. years)?
	° Represent the phenotype (e.g., therapeutic area, severity) of interest?
	• Is the size and complexity appropriate to the time and resources you can apply?
	• Is the biology represented appropriately?
	• Is the embedded biological knowledge current?
	• Is the original research context clear?
	• Are assumptions clearly stated?
	• Are assumptions appropriate for the new research context?
	• Are data and parameter sources appropriate for the new research context?
Uncertainty	• Does the publication identify key knowledge gaps and associated assumptions?
	• Does the publication evaluate the impact of key uncertainties via sensitivity analysis or “what if” scenario testing?
	• Does the publication include multiple Virtual Patients (VPs) to explore biological uncertainty that is relevant to the new research context?
Variability	• Does the publication identify known pathway variability?
	• Does the publication evaluate the impact of pathway variability via sensitivity analysis or “what if” scenario testing?
	• Does the publication comment on clinical variability?
	• Are multiple relevant VPs included?
	• If VPs are included, how do they differ from each other mechanistically?
	• If VPs are included, what clinical phenotype and response to therapy do they represent?
Testing	• Qualitative Testing:
	° Were relevant experts consulted to assess if model results looked reasonable?
	° Were relevant sources of information for qualitative testing identified and used, e.g., clinical data from related therapeutic areas, or relevant non-clinical data?
	° Were what-if experiments performed to assess model behavior?
	° Are subsystem behavior tests described, with appropriate data references?
	• Quantitative Testing:
	° Were relevant clinical data for the drug of interest used for testing?
	° Were relevant clinical data for drugs in the same therapeutic area used for testing?
	° Were multiple disparate types of model perturbations tested and compared to relevant data?
	° Did the model perform adequately, given the new research context?
	° Does the model include relevant clinical outcome measures and/or biomarkers?
	° Is it clear how the outcome measures were derived from the represented biology?
	° Were population-level outcomes reproduced with appropriate range and distribution of outcomes?