Identify putative mitochondrial and plastid targeting sequences in a protein sequence.
Highlights:
This is a neural-network-based prediction program capable of identifying ER signal peptides and mitochondrial or plastid transit peptides. In tests on a "virtual proteome" constructed from 6000 proteins of known location, GENOPLANTEx99 PREDOTAR achieves more than a 90% correct prediction rate overall.