Secondly, I’ve been thinking about this. There are several possible scenarios:
1.Depending on whether the splice mutation leads to exons skipping and frameshift, one would perhaps anticipate nonsense-mediated mRNA decay. So the mutated allele band could possibly be fainter, due to smaller size and NMD.
2. The mutation triggers the usage of a cryptic splice site close to the mutation in exon 21. In fact this is what is predicted by MutationTaster, and that prediction would lead to just 5bp being lost, frameshift and NMD would then also apply. If this happens, the band would not be resolvable on the gel.
I think the solution is that you should probably use proband cDNA, and sequence whatever bands are amplified. Hopefully then you’d see that there’s mixed signal onwards from somewhere near the mutation site, indicating a splice-induced insertion/deletion.