if you're not sufficiently sure with data given by NCBI, you should use UCSC (http://genome.ucsc.edu) and activate SIB track to get more idea on the relevance of the data. There are more tracks which can help you this way. After designing your primers, you can also use the primer design tool of UCSC and see directly in the genome browser if they work and the place of the amplicon. The only limitation is whether your species are in the database from UCSC.
It is not clear what "predicted using computational analysis" means. Most likely, the prediction refers to the interpretation of the nucleotide sequence as a sequence coding for a protein, not to prediction of the sequence itself. Thus, you don't need to worry about designing your primers based on the NCBI data base. Given the low price of oligos these days, you won't risk wasting more than a few $ (Rs) anyway