You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Annotate insertions at internal exon junctions as coding, not intergenic
A VCF insertion places its bases after the anchor, so an insertion anchored at
the last base of an internal coding exon lands at the exon-exon junction inside
the spliced CDS. The half-open exon interval used for overlap and per-segment
anchoring (< segment.end) excluded that anchor, so such insertions were dropped
from the transcript model and reported as intergenic.
Add Gene::insertion_at_internal_junction, which recognises an anchor on an
internal exon's last base (using transcript order: not the last segment on the
plus strand, not the first on the minus strand, so a CDS-terminal anchor stays
UTR). Use it in gene_overlaps_variant, variant_overlaps_coding_model, and the
per-segment insertion anchoring so the inserted bases are placed at the junction
of the reconstructed CDS and the protein effect is computed. The change is
additive: in-exon variants and every existing classification are unchanged.
Adds plus- and minus-strand regression tests (the minus-strand inserted bases
are reverse-complemented at the junction). 258 -> 260 lib tests; clippy, fmt and
the 36-scenario suite pass.
0 commit comments