Releases 2.5.0 and 2.4.x of MDAnalysis

We are happy to release the version 2.5.0 of MDAnalysis!

This is a minor update to the MDAnalysis library.

In line with NEP 29, the minimum required Python and NumPy versions have been raised to 3.9 and 1.21.0 respectively.

Supported Python versions: 3.9, 3.10, 3.11

Supported Operating Systems:

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

Note: this release also adds wheels for aarch64 and Apple M1 architectures for pip installations.

Upgrading to MDAnalysis version 2.5.0

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. Make sure you are using a Python version compatible with MDAnalysis before upgrading (Python >= 3.9).

Notable changes

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

Fixes:

• Fixed an issue where transformations were not being properly applied to Universes with multiple trajectories (i.e. using the ChainReader) (Issue #3657 #4008 PR #3906)
• Fixed an issue with the the heavy distance_type for WaterBridgeAnalysis where distance was not correctly assigned when more than one hydrogen was bonded to a heavy atom (Issue #4040, PR #4066).
• PDB topology parser no longer fails when encountering unknown formal charges and instead simply does not populate attribute (Issue #4027)
• Fixed an issue where using the between keyword of HydrogenBondAnalysis led to incorrect donor-atom distances being returned (PR #4092, Issue #4091)
• Fixed an issue where chi1_selections() ignored atom names CG1 OG OG1 SG and incorrectly returned None for amino acids CYS, ILE, SER, THR, VAL (Issue #4108, PR #4109)
• Fix H5MD reader to read box vectors rather than returning None as the dimensions (Issue #4075, PR #4076)
• Fix to allow reading NetCDF trajectories which do not have the time variable defined (Issue #4073, PR #4074)
• Allows shape_parameter and asphericity to yield per residue quantities (Issue #3002, PR #3905)
• Fix EDRReader failing when parsing single-frame EDR files (Issue #3999)
• Add ‘PairIJ Coeffs’ to the list of sections in LAMMPSParser.py (Issue #3336)
• PDBReader now defaults atom values for ts.data[‘occupancy’] to 0.0, rather than the previous default of 1.0. This now matches the default used when setting Universe Topology attributes using the first frame’s information (PR #3988)

Enhancements:

• ARM64 (osx and linux) wheels are now provided via PyPi (Issue #4054)
• Addition of a new analysis class analysis.atomicdistances.AtomicDistances to provide distances between two atom groups over a trajectory. (Issue #3654, PR #4105)
• Add kwarg n_frames to class method empty() in MDAnalysis.core.universe, enabling creation of a blank Universe with multiple frames (PR #4140)
• PDBReader now populates ts.data[‘tempfactor’] with the tempfactor for each atom for each frame. If an entry is missing for a given atom, this will default to a 1.0 value. Note, this does not affect the topology, i.e. AtomGroup.tempfactors is not dynamically updated. (Issue #3825, PR #3988)
• Add writing u.trajectory.ts.data[‘molecule_tag’] as molecule tags to LAMMPS data file (Issue #3548)
• Add progressbar_kwargs parameter to AnalysisBase.run method, allowing to modify description, position etc of tqdm progressbars. (PR #4085)
• Add a nojump transformation, which unwraps trajectories so that particle paths are continuous. (Issue #3703, PR #4031)
• Added AtomGroup TopologyAttr to calculate gyration moments (Issue #3904, PR #3905)
• Add support for TPR files produced by Gromacs 2023 (Issue #4047)
• Add distopia distance calculation library bindings as a selectable backend for calc_bonds in MDA.lib.distances. (Issue #3783, PR #3914)
• AuxReaders are now pickle-able and copy-able (Issue #1785, PR #3887)
• Add pickling support for Atom, Residue, Segment, ResidueGroup and SegmentGroup. (PR #3953)

Changes:

• As per NEP29 the minimum supported Python version has been raised to 3.9 and NumPy has been raised to 1.21 (note: in practice later versions of NumPy may be required depending on your architecture, operating system, or Python version) (PRs #4115 and #3983).
• Add progress bars to track the progress of mds.EinsteinMSD _conclude() methods (_conclude_simple() and _conclude_fft()) (Issue #4070, PR #4072)
• The deprecated direct indexing and times from the results attribute of analysis.nucleicacids’ NucPairDist and WatsonCrickDist classes has been removed. Please use the results.pair_distances and times attributes instead (Issue #3744)
• RDKitConverter changes (part of Issue #3996):
  • moved some variables (MONATOMIC_CATION_CHARGES and STANDARDIZATION_REACTIONS) out of the related functions to allow users fine tuning them if necessary.
  • changed the sorting of heavy atoms when inferring bond orders and charges: previously only based on the number of unpaired electrons, now based on this and the number of heavy atom neighbors.
  • use RDKit’s RunReactantInPlace for the standardization reactions, which should result in a significant speed improvement as we don’t need to use bespoke code to transfer atomic properties from the non-standardized mol to the standardized one.

Author statistics

This release was the work of 25 contributors, 15 of which are new contributors.

Our new contributors are:

@mglagolev @chrispfae @ooprathamm @MeetB7 @v-parmar @MoSchaeffler @xhgchen @DrDomenicoMarson @AHMED-salah00 @schlaicha @jvermaas @SophiaRuan @marinegor @g2707 @DanielJamesEvans

MDAnalysis 2.4.x

MDAnalysis 2.4.0 and three follow up bug fix releases were made earlier this year.

A full list of changes can be seen in the 2.4.x Changelog.

The 2.4.x branch was the work of 13 contributors, including the following 3 new contributors:

@jaclark5 @jfennick @Hakarishirenai

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