The format is based on Keep a Changelog, and this project adheres to Semantic Versioning.
- Will keep track of
TERin frames to correctly handle assigning chain IDs. - Check for combinatorial explosions.
- Modularized code and set up for deploying to PyPI.
- WISP now throws a more user-friendly error when it can't find a given atom in a given residue (e.g., when trying to calculate paths using CA, but some residues do not have CA atoms).
- Other minor stylistic updates the the code.
- Minor updates to deal with PDB files that have both ENDMDL and END tags.
- Formatting improvements to
README.md. - Changed development-branch name to
main.
- Reformatted code with Black Formatter.
- Updated email in copyright header.
- Updated
contact_map_distance_limitto 4.5 in example scripts, help descriptions, etc., to avoid confusion (given new default, see version 1.1 changes). - Added download and citation information to the standard output.
- Investigated Python2/3 differences. In brief, we recently noticed that WISP gives slightly different output when running under Python2 (numpy 1.16.0, networkx 2.2, scipy 1.2.0) vs. Python3 (numpy 1.17.4, networkx 2.2, scipy 1.3.1). We believe differences in the two numpy versions are responsible, though it is possible that similar differences may arise when running on different operating systems, etc. Regardless, the top-ranked paths are the same, but lower-ranked paths sometimes aren't. After some investigation and test updates to the code, we now believe subtle rounding-error differences to be the cause. For consistency's sake, we recommend using the same Python/OS/dependencies when comparing multiple WISP analyses.
- Changed the default
contact_map_distance_limitto 4.5 sowisp.pynow matchesWisp.tcl. - Updated the
README.mdfile. - Ensured Python2/3 compatibility.
- Modernized/cleaned up the code.
- The original version.