TODO

1. Revise coordinates written and read to file. Want these to be coordinates in the raw sequences (so if don't accept
edits there isn't a problem. Then need to be able to adjust these based on edits applied.

2. Don't load all consensus sequences into memory at once in apply
5. Create run routine which compares existing fasta to a new fasta and runs annotate/updates annotations for only
sequences which have been added or updated?
6. Passing an editfile into annotate.py will cause an error because consensus sequences can't be edited in place now
loading with index - handle
7. In annotate, improve edit finding so doesn't ADD FP edits which introduce more stop - obviously wrong


OLD
1. Want to be able to run all steps for just a specific region of the ref corresponding to an annotated feature
2. Handle rev complement features
3. Handle rev complement reads
4. Do we want to handle full codon deletions?
5. What is expected behaviour if have overlapping features and different edits found in each?