After proceeding with Prodigal for Genes prediction we would able to blast it and find known genes in our data. Is it possible to find novel genes by comparing it to conserved domain database CDD?
OK, so now I understand your main objective, which would be to discover new protein families by looking for short conserved domains from many currently established protein families, in your data.
In this case I still recommend using the Pfam database and DIAMOND to annotate and separate what fall into already described protein families from what could be new. Then, with the subset of ORF without representation in the Pfam database you can run RPSBLAST (ftp://ftp.ncbi.nih.gov/pub/mmdb/cdd/rpsbproc/) for your sequences over the CDD database (ftp://ftp.ncbi.nih.gov/pub/mmdb/cdd/cdd.tar.gz).