MDAnalysis 2.3 is out

04 Sep 2022

We are happy to release the version 2.3.0 of MDAnalysis!

This relatively small update to MDAnalysis reflects our commitment to doing more frequent (trimonthly) releases of the library.

In line with NEP 29, the minimum required NumPy version has been raised to 1.20.0 (1.21.0 for macosx-arm64).

Supported Python versions: 3.8, 3.9, 3.10

Supported Operating Systems:

Linux (on x86-64, aarch64, and ppc64le note: aarch64 and ppc64le can only be installed via pip for now)
Windows (on x86-64)
MacOS (on x86-64 and Apple M1)

Upgrading to MDAnalysis version 2.3.0

To update with conda from the conda-forge channel run

conda update -c conda-forge mdanalysis

To update from PyPi with pip run

pip install --upgrade MDAnalysis

For more help with installation see the installation instructions in the User Guide.

Notable changes

For a full list of changes, bugfixes and deprecations see the CHANGELOG.

Fixes:

Fixed reading error when dealing with corrupt PDB CONECT records, and an issue where MDAnalysis would write out unusable CONECT records with index>100000 (Issue #988).

Enhancements:

Formal charges are now read from PDB files and stored in a formalcharge attribute (PR #3755).
A new normalizing norm parameter for the InterRDF and InterRDF_s analysis methods (Issue #3687).
Improved Universe serialization performance (Issue #3721, PR #3710).

Changes:

To install optional packages for different file formats supported by MDAnalysis, use pip install ./package[extra_formats] (Issue #3701, PR #3711).

Deprecations:

The extra_requires target AMBER for pip install ./package[AMBER] will be removed in 2.4.0. Use extra_formats (Issue #3701, PR #3711).

CZI EOSS Performance Improvements:

A series of performance improvements to the MDAnalysis library’s backend have been made as per planned work under MDAnalysis’ Chan Zuckerberg Initiative EOSS4 grant. Further details about these will be provided in a future blog post.

MDAnalysis.lib.distances now accepts AtomGroups as well as NumPy arrays (PR #3730).
Timestep has been converted to a Cython Extension type (PR #3683).

Author statistics

This release was the work of 10 contributors, 3 of which are new contributors.

Our new contributors are: @miss77jun @rzhao271 @hsadia538

Acknowledgements

MDAnalysis thanks NumFOCUS for its continued support as our fiscal sponsor and the Chan Zuckerberg Initiative for supporting MDAnalysis under an EOSS4 award.

— The MDAnalysis Team

MDAnalysis