Open j-wags opened 3 years ago
Update: This does indeed seem to be a bug. Without control of the underlying machinery, however, it will be hard to fix these cases. We're going to prune these and move forward, assuming that the parameters that we would would harvest from these are covered by other structures.
Details:
David Cerutti 11:58 From what I could tell, the mol2 writer has a problem with determining bond patterns, order is another issue. The coordinates of the mol2 files are all in the right places, same as the corresponding PDBs I generated, but the bonds from atom to atom are all off, hence the H+ ions everywhere. But, we need Amber to give us a mol2, we cannot bypass this with some other software, right? Our duct-tape fix was incorporating Amber in this way to get around other issues. If the problem is this confined, to particular residues, then perhaps we can prune them and keep going with other residues providing coverage?
Jeffrey Wagner 13:56 Thanks for looking into this. We do need a mol2, but I agree that it's possible that these particular structures aren't essential to the FF port. We can prune them for now and revisit this if we start seeing issues in energy validation for these.
A few of the tripeptide mol2 files are malformed in some way. The backbone parameter file is winding up with some odd parameters with multiple charged
[H+]
s, indicating that a few mol2 files aren't being loaded correctly.I suspect that the following structures have the same problem:
NTerminal/THR_GLU
NTerminal/MET_TYR
NTerminal/LEU_MET
NTerminal/HID_MET
CTerminal_PRO_THR
I'll show NTerminal/THR_GLU (malformed) and NTerminal/ALA_GLU (correct) for comparison.
Looking at differences between the file text, it seems like atoms in the cap and first residue get shuffled in with each other. I can't tell if the atoms are valid-but-out-of-order, or if the bonding is nonsense as a result.
NTerminal/THR_GLU (malformed)
Loads into pymol like this:
With the body text below
NTerminal/THR_GLU/THR_GLU.mol2
``` @NTerminal/ALA_GLU (good)
With the body text below
NTerminal/ALA_GLU/ALA_GLU.mol2
``` @