Inversion mapping is not palindrome aware

[reece]

Originally reported by: Peter Causey-Freeman (Bitbucket: PeteCauseyFreeman, GitHub: Unknown)

---

Hi All,

I'm reporting this as a bug but nothing is actually broken.

It may be a fussy pedantic thing, but the mapper for hgvs is not palindrome aware when mapping inversions. For example, I queried both hgvs and mutalyzer with the variant description NM_000088.3:c.589_600inv into both mutalyzer and hgvs.

hgvs returns (workflow evm.c_to_g then evm.g_to_c to collect sequence data) NM_000088.3:c.589_600invGGTCCCCAAGGC

mutalyzer, however returns the following message (NOTE, mutalyzer reports the :n. coordinates not the input :c. coordinates in its warning message)

"Sequence "GGTCCCCAAGGC" at position 715_726 is a partial palindrome (the first 1 nucleotide(s) are the reverse complement of the last one(s)), the HGVS notation prescribes that it should be "GTCCCCAAGG" at position 716_725."

Although the inversion generated by hgvs was not technically incorrect and the protein consequence is the same as reported by mutalyzer (I have written some code to generate protein consequences from hgvs nucleotide inversion objects), HGVS nomenclature would dictate that the correct output would be NM_000088.3:c.590_599invGTCCCCAAGG not NM_000088.3:c.589_600invGGTCCCCAAGGC.

Regards

Pete

---

• Bitbucket: https://bitbucket.org/biocommons/hgvs/issue/285

[reece]

Original comment by Reece Hart (Bitbucket: reece, GitHub: reece):

---

fixes #285, #334, #335, #324, #340: inversions parsing, formatting, and normalization