When splicing is not all or none: GT>GC 5′ splice-site variants as a model for intermediate effects and challenges in variant classification

Variants with intermediate functional effects—those that neither eliminate nor fully maintain normal gene function—represent an underrecognized source of genetic complexity and constitute a grey zone that challenges variant classification. Here, we address this issue using GT>GC (+2T>C) 5′ splice-site variants as a tractable model, as approximately 15% of such substitutions retain variable amounts of wild-type (WT) transcript. Using residual WT transcript as a quantitative readout, we first show that disease-associated GT>GC variants known to generate substantial WT transcript (for example, SPINK1 c.194+2T>C, HBB c.315+2T>C, and BRCA2 c.8331+2T>C) consistently complicate classification efforts. We then carried out a locus-wide assessment of all 26 GT>GC substitutions in CFTR , applying SpliceAI delta donor-loss scores as an initial screen and cross-checking classifications in expert-curated database. This evaluation indicated that three variants with low or intermediate SpliceAI scores are likely to retain WT transcript. Minigene analyses of four selected CFTR variants, together with additional examination of a BAP1 GT>GC variant with conflicting clinical interpretations, illustrate the methodological challenges in predicting and experimentally quantifying residual WT transcript. These results emphasize that (i) in silico predictors cannot reliably discriminate between complete and partial splice disruption and (ii) experimental systems differ in their ability to detect subtle splicing outcomes. Overall, our integrative analysis shows that GT>GC variants generating appreciable WT transcript exemplify a broader class of intermediate-effect alleles that complicate both computational prediction and experimental assessment, underscoring the need for classification frameworks that incorporate quantitative functional data and better capture the continuum of variant effects.