v0.3.1: improved user experience

1. prepare_dataset.sh script was improved to provide an easy option to tell the script what segments you want to analyze and whether you want to use IQ-Tree to build trees (for precision reassortment analysis) or FastTree (for quick analysis on large data). See the tutorial for more details.
2. More robust reassortment rate estimation: TreeSort now outputs two reassortment rate estimates (Bernoulli and Poisson). We recommend using the first estimate for less dense datasets and the latter for larger and very dense datasets.
3. TreeSort now robustly labels the internal nodes of the annotated output tree file with TS_XXX labels.
4. Other minor improvements and bug fixes.