For instance, RPE65 mutations can be missense frameshift etc, but is there an efficient process to know whether the vast majority of mutations lead to a functional loss?
It may help you to look at our recent paper on single-site mutations – variants of unknown significance – where we used structure (which could be an AlphaFold model) and sequence conservation in terms of evolutionary relationships.
See Front. Mol. Biosci., 14 December 2021 | https://doi.org/10.3389/fmolb.2021.791792