From karplus@cse.ucsc.edu Fri Jun 9 13:33:33 2000 Return-Path: Date: Fri, 9 Jun 2000 13:33:30 -0700 From: Kevin Karplus To: karplus@cse.ucsc.edu, rph@cse.ucsc.edu, saira@sanger.ac.uk, cbarrett@cse.ucsc.edu, cline@cse.ucsc.edu, markd@cse.ucsc.edu, cbarrett@cse.ucsc.edu, stu@cse.ucsc.edu CC: leslie@cse.ucsc.edu Subject: prediction for N-domain of T96 I have what I think is a pretty good prediction for domain 1 of T0096: In pce/casp4/t96 see-a2m 1bi0/1bi0-T0096-fssp-global.pw.a2m.gz The SAE problems in handling sequences which don't correspond to the residues that occur in the ATOM records of the PDB file cause some misalignment in the picture, which is a bit frustrating. That is why the 1bi0 sequence occurs in the alignment twice. The shorter second sequence corresponds better to the ATOM list (it may not be perfect), so using it as the guide sequence gets the alignment right. The second domain seems harder, and we may have to search for it separately. The T99 server for CAFASP was given a first and second domain (split a bit differently than I would have predicted), and predicted 1b78A in both directions (through rather weakly). I haven't built alignments for the second domain yet, and an all-protein alignment with 1b78A did not have a domain boundary near where we want one.