v0.2.5

• Update the sampling for Var_e, it sometimes breaks if Meta-ed data has some errors.
  (Don't need to re-run, if your results make sense in previous versions. However, if the median Var_e is very small (< 0.5), a re-run is suggested)