Please rephrase this question as complete sentences. It is not clear at all what you want to do, please provide additional information about your project. In principle you can use any blast database, obviously which one you choose depends on your question.

Saw your edit, still got some questions:

Possibly your question is to identify pan/core-genomes of bacterial species?

• are all or some of these sequences already in the database, or are all or at least some of them newly sequenced?
• How many comparisons do you want to make (hope not all bacteria in NR vs. all)?
• Do you want to use a restricted subset based on tax-ids (e.g. all Bacteroidetes)?
• How closely related are those? (determins whether to use nucleotide or AA sequences)

In any case you can most likely (if you need more than a few bacterial genomes) start to download the complete NR and NT databases, those can then be filtered by taxon using the NCBI gi<->taxid mapping file.