Megamash cannot call complicated sequences

`seq_recovery=0.4152_0` for example, cannot be processed into a megamash table because it doesn't have any unique 16mers. This is because of a flawed assumption: ESPECIALLY in protein variant libraries, 16mer windows won't necessarily be unique - and even if a sequence is unique, a 16mer sliding window may not be able to pick it up. This is a real issue.

I'm still trying to figure out how to fix this.

[templateMap.csv](https://github.com/Koeng101/dnadesign/files/14300542/templateMap.csv)

Start by reproducing the megamash failure with seq_recovery=0.4152_0 and inspect the attached templateMap.csv. Trace how sequences are converted into a megamash table and identify where the unique-16mer assumption rejects this input. Done means complicated protein variant sequences can be processed without requiring unique 16mers.

Written by the indexing model from the issue text.