normal Martinize: Charged/polar side chains at termini

  • rvanlehn
  • rvanlehn's Avatar Topic Author
  • Offline
  • Fresh Boarder
More
8 years 2 months ago #5370 by rvanlehn
I was having trouble with LINCS crashes when switching a protein from a MARTINI v2.2 representation to v2.2P, using the polarizable water / polarizable side chains. I just found the error and thought I'd report it for the benefit of others.

The problem arises when using the default settings of martinize.py, which sets the backbone beads of the N- and C-termini to be charged (either a Qa/Qd bead type with a +/- 1 charge). In v2.2P, polar and charged side chains have dummy are free to rotate but are constrained to either each other (for polar residues) or the side chain (for charged ones). When the N- or C-termini are polar/charged residues, I found that the SC beads would rotate then move toward the BB due to the electrostatic attraction until eventually the system blew up with LINCS errors. This specifically happened for a terminal THR and terminal GLU. I didn't observe this for v2.2, likely because the SC beads are treated differently and the dielectric constant is higher. The problem was fixed by either manually modifying the backbone bead types/charge, or using the -nt option in martinize.

There isn't a bug, but I thought I'd report this issue in case other users are having a similar problem; alternatively, perhaps martinize can print a warning for this specific circumstance. Please let me know if there is also something potentially wrong in my particular workflow - I use Gromacs 5.0 with the recommended MARTINI MDP options for v2.2P.

Best,
Reid

Please Log in or Create an account to join the conversation.

More
8 years 1 month ago #5381 by peterkroon
Replied by peterkroon on topic Martinize: Charged/polar side chains at termini
Hi,
Thanks for the report and sorry for the late reply.

For completeness' sake, what water model are you using, and could you post the complete contents of your mdp file? Just to make sure we're looking at the same thing.

Peter

Please Log in or Create an account to join the conversation.

Time to create page: 0.087 seconds