The role of transcript context in evaluating splice-region variants
Splice-region variants have always been tricky to evaluate, especially those that aren’t located exactly at the canonical donor or acceptor sites but still close enough to possibly interfere with splicing. One of the biggest challenges we face in variant interpretation meetings is understanding how such variants might affect alternative isoforms, especially when multiple transcripts exist for the same gene. Without a clear visual of the transcript landscape, it’s easy to overlook important exon-intron boundaries or misjudge whether a variant truly disrupts the most biologically relevant isoform. I'm curious how other groups here handle this type of analysis. Do you rely solely on prediction algorithms like SpliceAI, or are you integrating visual tools into your workflow? And if so, which ones actually help clarify the transcript-specific consequences in a way that’s practical for collaborative review?
For me, visual interpretation has been a game changer. I recently started using this tool: https://compassbioinfo.com/. It’s helped me analyze splice-region variants in much greater detail by showing exactly where a variant sits in relation to each transcript isoform. I’ve had several cases where what looked like a benign intronic variant actually overlapped a cryptic exon in an alternative transcript — something that would’ve been missed using standard annotations alone. This kind of visualization also helps me explain the reasoning behind my classification to clinicians and lab staff during variant review meetings. Plus, it lets us quickly compare transcript structures across genes, which is critical when dealing with complex splicing scenarios. Overall, it’s made my interpretations more confident and less reliant on assumptions.