Hi The two "fragments" are covalently linked (so it is the first case), they are not actual fragments but a deletion mutant, the aim was to transfer my assignments from the original chain onto a chain where the residues were deleted. I also wanted to keep the numbering consistent between them both. I will try it again with the python code you gave. However, with the two fragments, I got Copy Chain Assignments to work, I manually inputted the Residue Mapping by setting the destination for each residue. Copy assignment seemed to work fine in this instance and I got no errors and afterwards the resonances looked sensible and my Assignment Graph looked OK as well going from residue 136 to residue 169 directly i.e. 134 135 136 169 170 171 (the 50residues and 2deleted was just to help describe the situation!). So it looks like it has all worked, but if I understand correctly then it possibly hasnt. Would I be better to back track and repeat it with one chain and then renumber the 2nd half of the chain? Or should i be fine to carry on now it has (apparently) worked? It sounds like I may have two fragments uncovalently linked at the minute which may cause problems? Cheers Ben -----Original Message----- From: CcpNmr software mailing list on behalf of Wayne Boucher Sent: Thu 26/04/2007 4:56 PM To: [log in to unmask] Subject: Re: Experiments linked to Molecular Systems Hello, Right, I think we need to go back one step first. Does the molecule with the two-residue gap have the two residues at either side of the gap as "middle" residues (so are covalently connected or not covalently connected but for some reason you want to pretend they are) or is the "top" part of the gap really a C-terminus and the "bottom" part an N-terminus? In the first case, the data model says that you should create one standard molecule, so you should *not* create those two sections the way Tim wrote in a previous email. That should only be done in the second case. When you add the molecule template to the molecular system, in both cases you end up with one chain but in the first case that chain has one (so-called) chain fragment and in the second case the chain has two chain fragments. In the first case the molecule (template) is what is called "standard linear" and in the second case it is not. In API terms, there is an attribute, molecule.isStdLinear, which is True in the first case and False in the second. There is another attribute, molecule.stdSeqString, which gives the "standard sequence". But if the molecule is not standard linear then that standard sequence is just the empty string. So in the first case the standard sequence is what you would expect but in the second case it is just the empty string. Perhaps this standard sequence algorithm needs some souping up to cope with these "almost standard linear" situations. The relevance to the "copy chain assignments" dialog is that to get the residue mapping, the standard sequence of the two chains is compared using some sequence alignment algorithm. This is going to fail completely for the not standard linear case because the sequence is empty. This explains what you have just observed. If your molecular system is really the first case (so the "gap" is not really therer) then currently the Analysis graphical interface doesn't let you change some of the seqCodes by +2 but it will let you do the "copy assignments". To change the seqCodes you would have to run that bit of Python script I gave. It should not be that difficult (but more than five minutes) to give you a way of accomplishing the same thing via the graphical interface. If your molecular system is really the second case, then the Analysis graphical interface lets you change the seqCodes, as you already tried, but then the "copy assignments" fails because the sequence alignment fails because of the reasons given above. It will be less trivial to sort this problem out than in the other case. Wayne On Thu, 26 Apr 2007, Ben Goult wrote: > Hi > > Sorry to keep posting but I have nearly got this to work now. I made the > chain with the two fragments and now want to use copy chain assignments onto > this new chain. > > However, when I select this chain as the destination chain then the residue > mapping does not work. It does not predict any destination residues for the > mapping. I do seem to be able to manually select the destination for each > residue. > > If I make a 3rd chain with the identical sequence but consisting of only one > fragment (and thus different numbering for the second half) then the Copy > Chain Assignments works fine. Therefore it seems that perhaps the Copy Chain > Assignments does not like fragmented chains? > > Cheers > > Ben >