openvax / topiary / 24404280509

88%

master: 90%

Build Type

Pull #121

github

Committed by iskandr

Commit Message

Add frameshift fragment regression tests — v5.4.0

Pin _fragment_from_effect behavior on varcode FrameShift /
FrameShiftTruncation effects, addressing the PR B review concern that
inframe=True would drop downstream novel 9-mers under
only_novel_epitopes. The 75-case suite verifies the varcode invariant
(aa_mutation_end_offset == len(mutant_protein_sequence) for FrameShift),
that inframe=True and inframe=False produce identical target_intervals
for frameshifts, per-peptide overlaps_target across lengths 8–11, and
that only_novel_epitopes preserves every downstream 9-mer.

Pull Request Pull Request #121: Refactor predict_from_variants onto AntigenFragment (PR B)

Coverage Stats

94 of 98 new or added lines in 4 files covered. (95.92%)

5 existing lines in 1 file now uncovered.

3004 of 3409 relevant lines covered (88.12%)

2.64 hits per line