Normally, when we are searching with a certain sequence we are most concerned with thos sequence of the of the same type. I.e. if we enter a protein sequence, we will want to find other similar protein sequences.
However, as nucleic and protein sequences can be translated back and forth, if we want to be confident in our prior art search, it is worth running both types of searches.
If we are doing a landscaping of a technology space and we are prioritizing speed of results, this double search may not be necessary. However, if the purpose of our search is for full Freedom to Operate (FTO), it may be worth investing the extra time.