search

glygener / glygen-issues

Repository for public GlyGen tickets
GNU General Public License v3.0
0 stars 0 forks source link

Disconnected sulfate in 3D structure #338

Open ReneRanzinger opened 1 year ago

ReneRanzinger commented 1 year ago

Example: G13157WZ

If you rotate the example 90' up you will see that the part with the yellow node connected to 3 red dots (SO3 aka sulfate) is disconnected from the structure. There is another red dot close by that it is supposed to connect. There is no issue in the PDB file. Its Mol* that does not make the connection.

Can you contact Dan or however you worked in Robs group to see if there is a setting we have to make (similar to the one to show the SNFG shapes). If they can not help please report it in the Mol* github. Attach an screenshot and the PDB file.

sujeetvkulkarni commented 1 year ago

there should be bond between sulphur and oxygen but its not connected in Mol*.

sujeetvkulkarni commented 1 year ago

I see this behaviour in Mol* and UCSF Chimera. It seems when drawing bonds they look internally within a residue and then only sequentially connected neighboring residues. If I shift the sulfate residue up a residue in order then the sulfate is bonded, but the carb residues aren't.

A potential fix is if we add a CONECT card (tells it to bond certain atoms). This works for Chimera, can you check it with MolStar (attached)? I'm unable to upload files directly to their online viewer.

If this works then you'd have to get all the files again from us with the CONECT cards included. Perhaps there is a way to get MolStar to check for bonds between each residue, which would be much less effort. I've asked on the MolStar page (https://github.com/molstar/molstar/discussions/823

Bond by distance beyond sequentially connected residues in PDB files? · molstar/molstar · Discussion #823 When running MolStar on our website we have branched molecules which results in some of them not having bonds drawn (see the sulfate here): There should be a bond between the yellow sulfate atom an... github.com ). I see you didn't get a response to your earlier post, did you figure it out or shall I email the developer directly? We've worked on stuff before and he's been responsive.

sujeetvkulkarni commented 1 year ago

@ReneRanzinger please see below Oliver's response. He has implemented the pdb file changes.

Any luck Sujeet?

If it's not possible to overcome this, you can build everything again. I've updated the output pdb files to contain the necessary CONECT cards, so the dynamic bonds fix won't be necessary.

Best, Oliver Grant Research Scientist - Woods Lab Complex Carbohydrate Research Center University of Georgia

ReneRanzinger commented 1 year ago

@sujeetvkulkarni OK, we want to rerun the 3D structure generation in this release anyway. I will assign the ticket to me and we test it afterwards.