I am working on annotating a few hundred transcripts that may or may not be novel. It is likely that most are orthologs that I was unable to detect via genome alignment, while a couple may be truly novel or have a more distant ortholog. To approach this, I have been running blastp locally against the nr database. However, this is incredibly slow. I need something faster. What do people recommend for this? I looked at OrthoMCL, but that requires having a mySQL database that I can write to.