Q: What makes exTSP different from MutPred2 or AlphaMissense?
MutPred2 and AlphaMissense provide tissue-agnostic estimates of functional impact. exTSP adds tissue and isoform context by integrating transcript-level expression, enabling prioritization of variants affecting transcripts preferentially expressed in relevant tissues.

Q: Does exTSP work for non-missense variants?
This server is intended for missense variants. Noncoding, splice-site, frameshift, and nonsense variants are not the primary target of this model.

Q: Why do I see different scores across tissues?
Scores change because exTSP explicitly models tissue relevance through transcript expression. A variant may be more relevant in one tissue if the impacted transcript is preferentially expressed there.

Q: Why might the top transcript differ from MANE Select?
MANE Select is a standardized clinical reference, but disease mechanisms may involve alternative isoforms. exTSP prioritizes isoforms based on tissue-specific expression patterns.

Q: Can exTSP be used to “call” pathogenic vs benign?
exTSP is best used for ranking and evidence generation. Threshold-based classification depends on the application, dataset, and desired false-positive rate.

Q: What expression resources are used?
exTSP uses transcript-level expression from GTEx (adult tissues) and BrainSpan (developmental brain).