next up previous contents index
Next: Choosing burn-in Up: Documentation for TreeLD, version Previous: Exporting images   Contents   Index


Choosing parameters for the Markov Chain Monte Carlo

Obtaining reliable results from TreeLD depends on multiple factors. Some, like the informativeness of the marker map cannot be influenced. On the other hand, the burn-in of the MCMC, the number of trees generated and the number of focal points chosen are important parameters for the analysis of a dataset that the user assigns. While the optimal value for each parameter depends on the individual dataset, the following describes some considerations for a reasonable starting point for an analysis.



Subsections

Sebastian Zoellner 2005-01-27