I have obtained nucleotide and protein genes using PRODIGAL tool from assembled contigs in a illumina sequenced metagenomic sample. I need to search against three databases for antibiotic resistant genes (ARGs) which are CARD, ResFinder and NDARO.
CARD and NDARO have both protein and nucleotide sequences and ResFinder has only nucleotide sequences. Also one of the query reads show similar hits for two ARGs (lsaE in CARD and lnuB in the other two databases which have identical sequences but different name and annotation).
Can I use BLASTn for ResFinder and BLASTp for the other two databases and merge the results based on all unique query reads or do BLASTn for a combined non-redundant database?
While using BLASTp I get more number of my query genes aligned than in BLASTn while using the same parameters (For the first method using BLASTp and BLASTn - around 380 queries aligned and using BLASTn alone against combined database - around 230 queries aligned).
Which of the methods can be done? Please help?
