New year, new release!
Changes
- Updated and polished the documentation
- By default (and if installed) martinize2 will use the mdtraj module to determine protein secondary structures, instead of requiring users to install the exact right version of dssp
- Revamp how martinize2 deals with the Go model. Martinize2 can now parse and process the rCSU contact maps which means users no longer need to run another script to produce a functional topology! Thanks @Lp0lp @fgrunewald !
- Users can specify desired modifications via the command line, but mistakes are easily made. Thanks to @csbrasnett we now produce a friendly warning in these cases.
@csbrasnett found and fixed an issue with produced elastic networks in some rare cases. See #559 for more details, - @Ladme found an issue with Martini 2.2 topolgies where links override parameters from modifications. In particular, the result was that terminal residues with secondary structure T, E, F, H, 1, 2, or 3; or terminal ALA, PRO, or HYP residues with secondary structure S would end up with a charged BB bead with beadtype N0 rather than the appropriate Q type. This is now fixed by filtering when and where Links get applied based on the modifications present.
- Small maintenance
What's Changed
- Replace deprecated pkg_resources with new importlib by @pckroon in #546
- Make sure DATA_PATH is always a pathlib.Path by @pckroon in #549
- Cleanup to Go implementation by @Lp0lp in #555
- add warning when mod not found by @csbrasnett in #556
- -el fix by @csbrasnett in #559
- Documentation by @pckroon in #545
- Use MDTraj dssp by @pckroon in #547
- Implementation of the Martini3 Go-model by @fgrunewald in #550
- Issue 560 by @pckroon in #561
- Log a helpful message when an element is not present in the ATOM_MASS dict by @pckroon in #571
Full Changelog: v0.9.6...v0.10.0
Contributors
pckroon, fgrunewald, and 3 other contributors