Hi Simon, Thanks for the pointers. My nt database is segmented into 512 MB chunks. I have found that the NCBI "fastacmd" command will do for now. I do, however, have one more question: Does anyone know how one would go about getting output which contains only a list of accession numbers of the matching sequeces from running a Blastp? Thanks, Nick > By any chance is the file which contains U51243 > 2Gb in size? The > current version of dbiflat fails to index any sequences past the 2Gb > limit in any file and just silently drops them. > > There is a fix in the works in the next EMBOSS release, but for now the > only fix is to split your database files into <2Gb chunks. > > If this isn't it, try running seqret -debug nt:U51243 and look at the > seqret.dbg file for clues as to what's going wrong. > > Simon.