**SNAPP for inferring species trees and species demographics from SNP data**
Here we provide the input files and outputs of two independent SNAPP analysis runs. We also provide a combined tree file from both runs, which was used to generate the associated maximum clade credibility tree.
We investigated the evolutionary relationships among populations using the SNP data in a coalescent framework with SNAPP v1.3 (Bryant et al. 2012) implemented in BEAST2 v2.4 (Bouckaert et al. 2014). The SNAPP model is based on the coalescent process and therefore assumes that shared polymorphisms among lineages are due to incomplete lineage sorting and not gene flow (Bryant et al. 2012). We subsampled each distinct population (determined by DAPC and clustering analyses) to include 2–6 representatives, for a total of 19 samples, estimated mutation rates (u and v) from the data (0.997 and 1.002, respectively), and fixed the birth rate (λ) of the Yule prior to 25. We performed two independent runs with a chain length of one million generations, sampling every thousand generations. We examined convergence using Tracer version 1.6 (Rambaut & Drummond 2009) and created a maximum clade credibility tree from the post-burn-in sample. To obtain a rough estimate of divergence dates we converted branch lengths using a human mutation rate of 1x10-8 (Lynch 2010), as no estimate currently exists for amphibians.
