Team 3

Use-cases and power-testing for integrating simulation with popgen analysis

Members:

• Allan Strand (team lead)
• Michelle DePrenger-Levin
• Sean Hoban
• Eric Archer
• Libby Liggins
• George Shirreff
• Christian Parobek

Goals

Create an R package that guides users through the workflow of implementing simulations in population genetics questions. Create accompanying documentation that outlines the questions that simulations can address and discusses issues relevant to choice of initial parameters, simulator type, execution, and summary metrics.

Products

Status:

1. Choose broad type of problem (1 of Power analysis, Performance Analysis, Null Models, [maybe abc])
2. Gather information from user input (both species and markers system)
3. Based on 1 and 2 offer advice on the best simulator
4. With user input, gather parameter values by:
5. querying the user (text or shiny)
6. querying an uploaded dataset
7. Special parameter: the simulation summary functions required to address the problem above
8. construct the actual simulation inputs (no user input)
9. run simulation rep
10. analyze simulation rep with the simulation summary functions outlined above, based on a parameter choice, save the simulations or discard them
11. assemble the summaries of the simulation reps into a 2 element list
12. collection of the parameters driving the simulations (probably a data frame)
13. collection of the output summary statistics (almost certainly a data frame)
14. Summarize the data frame assembled in step 8.  The summarization function will depend on choices in step 1.  At a minimum, we will summarize power analyses, performance analyses, and null models.

• products go here