Rappsilber-Laboratory / build-xiview

GNU General Public License v3.0
2 stars 0 forks source link

discrepancy output files mapping onto homomers #115

Open giulia9picco opened 5 days ago

giulia9picco commented 5 days ago

Good afternoon, I've been using the 3D NGL export to retrieve the cross-link distances and pseudobonds for provided homomer structures (AlphaFold predictions), in order to have, among all possible combinations of ambiguous intra-links, only the ones showing the shortest distance. After some time plotting and visualizing the cross-links retrieved in this way, I have realized that there is a discrepancy between the cross-links mapped via the 3D NGL (and their downloaded files), and the file downloaded via 'export - filtered crosslinks', after selecting the protein corresponding to the uploaded structure from the network. A short example : I have a network generated with my cross-link dataset, I upload to xiVIEW a structure for a homotetramer of one of the proteins in the network. If I visualize the structure via NGL ('show - shortest possible cross-links only' is selected, but nothing changes if I do not select it) I see and download something like :

while, exporting the file from the main page ''export - filtered crosslinks", gives from the same cross-link:

This happens with most of the ambiguous intra-links, sometimes are mapped within the same subunit but the subunits are different in the two cases (A-A in file 1, B-B in file 2), resulting in a difference distance, or with intra-links categorized as inter-links but connecting different pairs of subunits depending on the file downloaded (A-B in file 1, B-D in file 2). By visualizing all possible combinations on Chimera, looks like the shortest combinations are actually present in the "export - filtered crosslinks" file. Do you know what could be the reason for this discrepancy ? And, if the correct file is indeed the one provided from the main page (and not NGL), how can I retrieve the same sort of files that NGL provides, as pseudobonds, but with the correct cross-link combinations?

Thanks in advance for your support Best, Giulia

colin-combe commented 4 days ago

Hi Giulia, could you share the url of the data upload and the 3d structure file you uploaded with me, please. My email address is colin.combe@ed.ac.uk if you prefer to keep it private, cheers, c