Minimal-Shared Regions Refining

RNA sequencing has revolutionized gene expression analysis, but accurate transcript quantification remains challenging due to ambiguities in read assignment. These ambiguities often stem from overlapping exonic regions shared among homologous transcript isoforms. Our analysis shows that these extensive, undefined shared regions-sequences that overlap across exons or transcripts of both homologous and heterologous genes-are prevalent across multiple species, including humans and common model organisms. To address this, we developed a minimal-sharing strategy that segments each transcript into unique sets of k-mers. These k-mers are systematically compared across all transcripts to assess and quantify overlap. This approach isolates minimally shared regions specific to each isoform, effectively resolving complex overlaps and significantly reducing mapping uncertainty.