Null variants in 3'-end of gene
Null variants (e.g. frameshift, stop_gained, +/-2bp splice) that result in a truncated protein may not be important if they occur in the last part of the gene, generally involving the last exon. For some (typically very well characterised) genes, there may also be other, specific boundaries (a particular codon) where mutations occuring downstream of that boundary have shown no discernable effect. This is valuable information particularly when evaluating the ACMG PVS1 tag.
Ideally, setting specific boundaries should be possible for individual genes (relates to #150), but this would then have to be automated with a trusted source. A simpler, temporary solution was implemented in LA-650 (flag last exon).
A special case involves the possibility for non-sense mediated decay (NMD) for stop gain variants, see #22. A solution could possibly involve both that and this issue.