fix: return unknown when initiation site is affected

[nokara26]

I fixed N-terminal extension and to return unknown by default when initiation site is affected.
p.M1*ext-? is not valid because changing to Ter is incorrect.
And I found N-terminal extension is a variant in the 5' UTR and it is difficult to predict accurate hgvs.
So I fixed to return unknown by default when initiation site is affected.

https://hgvs-nomenclature.org/stable/recommendations/protein/extension/

p.Met1ext-5
a variant in the 5' UTR activates a new upstream translation initiation site starting with amino acid Met-5.
NOTE: modified from p.Met1extMet-5.

p.Met1_Leu2insArgSerThrVal
amino acid Met1 is changed to Val, activating an upstream translation initiation site at position -4 (Met-4), inserting amino acids ArgSerThrVal between Met1 and Leu2.
NOTE: this variant is not described as an extension (p.Met1Valext-4) since Met1, part of the normal amino acid sequence, is changed.

[nokara26]

to use prefer-extension-for-initial-codon-alt? in protin-extension because protein-extension is called from protein-indel.

[nokara26]

to move one position to the N terminal side.

[federkasten]

Sorry for the delayed reply 🙇 LGTM!