Isoform-level characterization and structural diversity of antithrombotic genes in Hirudinaria manillensis.

Antithrombotic proteins of medicinal leeches-originally characterized from salivary-gland secretions-suppress coagulation, platelet activation and fibrinolysis at the feeding site, but the contribution of alternative transcript processing to this repertoire remains poorly documented. Here we used four H. manillensis RNA-seq libraries (THA-Hm1-THA-Hm4), whose species identity and overall sequencing performance were characterized in our companion study, to determine how many antithrombotic loci produce multiple structurally distinct transcript isoforms and how such variation extends into coding capacity. Reference-guided assembly merged into a unified transcript set identified 38 transcript-supported candidates in addition to the 72 catalogued reference genes. When candidate transcripts were grouped by parent MSTRG locus and ranked by isoform count, six loci carried three or four structurally distinct models: MSTRG.4545, MSTRG.5212, MSTRG.3310, MSTRG.5709, MSTRG.5200 and MSTRG.5713. Transcript-length variation alone produced isoforms at three of the six loci, whereas at MSTRG.5200, MSTRG.5709 and MSTRG.5713 the variation extended into the coding sequence, producing predicted proteins ranging from 49 to 1159 amino acids. These results reveal localized isoform complexity at a small set of antithrombotic loci and define a focused candidate set for downstream domain annotation, splice-junction visualisation and functional validation.