Badread v0.2.0

I made this version to include new Nanopore error/qscore models. These are named nanopore2020 and are now the default. For consistency, I renamed the old models from nanopore and pacbio to nanopore2018 and pacbio2016.

I also increased the default read accuracy by a bit from --identity 85,95,5 to --identity 87.5,97.5,5. This is to account for the fact that long reads are more accurate today than they were a few years ago. I still wanted Badread's defaults to be worse than real reads, but not too much worse 😄