This dataset provides ready-to-use protein sequence databases for 20 distinct biomes. Each database was constructed using genomes from the 20 most abundant genera identified within the corresponding biome. Biome-specific protein databases were created using open-access 16S rRNA gene-targeted amplicon sequencing data. A total of 26,293 open-access samples were analyzed to identify 20 most abundant bacterial genera for 20 different biomes. 299,356 genome and 145,662,299 protein sequences for these genera were compiled to create ready-to-use databases.