Dear all
First, I am not sure whether I should direct this issue to aria or ccpnmr mailing list ... (so l do both)
I am encountering an issue that I do not understand at all.
After having finished my structure calculation using aria interfaced to a ccpnmr project I started to fill in the form for database submission (ECI) and, as usual, had to decide whether to use the residue numbering of my protein construct or that of the published protein sequence. My ccpnmr project that was used within aria followed my personal numbering. I know that I can ask BMRB or PD staff to change - this is not my issue here.
I decided to try changing residue numbers by applying a correction via the "Molecule" menu - consequently, this changed all apparent numbers within my resonances, peak lists etc. Then I changed numbering of the talos.tbl file and started structure calculation using exactly the same noesy spectra as before (started from my successful structure calculation ccpnmr project saved with a different name before changing residue numbers and after having previous structure and restraint lists as well as aria derived peak lists removed) and the same aria project in which only the paths to the ccpnmr project and talos file were changed.
What happens is that all long=range connectivities get lost during the structure calculation process and the number of NOEs used (aria report file) decreases from one iteration to the other.
However, the noesy peak lists up to it8 (stored aria xml files, or peak lists created by aria in the associated ccpnmr project) are correct and contain the same number of assigned noe peaks than within the reference structure calculation. Assignments in teh created peak lists are correct and are correctly numbered. But it seems that they are not used anymore for structure calculation.
How could I check what happens ? Any clues what is going wrong ? I suppose that the ccpnmr project contains some hidden links to the previous residue numbering, but I do not understand where.
The aria Cache parameter was set to "no".
Thank you
Beate
|