Due to changes to `gumpy`, this had to be updated to match.
When a synonymous mutation is found, there may be cases in which it is important to know specifically which nucleotides in the codon changed. For example, in the WHO catalogue, `fabG1L203L` has three different predictions depending on which nucleotides are changed:
If the change is `fabG1g609a`, it confers `R`
If the change is `fabG1c607t`, it confers `U`
If the change is anything else, it should confer `S` due to the default rule `gene*=` confers `S`
Therefore an update was made to include the nucleotide changes within each synonymous mutation. Within the WHO catalogue, this takes the form of all synonymous mutations being converted to multi-mutations with their nucleotide variants. This allows the above to be represented within single rows:
`fabG1L203L&fabG1g609a` confers `R`
`fabG1L203L&fabG1c607t` confers `U`
`fabG1L203L` confers `S`
`gumpy` was then updated to match this logic - pulling out such nucleotide variants for all synonymous mutations.
Within `gnomonicus` then, this results in nucleotide variants being mixed into the MUTATIONS in such cases, but these are excluded from the EFFECTS in every case other than when the associated multi-mutation lies within the catalogue. The result is that the nucleotide is never considered as a stand-alone mutation for the sake of EFFECTS, but is included as part of a multi-mutation iff the catalogue specifies this.