back-map protein mutations to genomic mutations

we have many protein mutations for which we don't have the corresponding genomic mutations from databases

it would be possible to back-map the protein mutations we have at genomic level, i.e. identify those possible genomic mutations (at least as SNVs) that result in the protein mutations