BioEmu 1.2.0 includes support for extra residue embeddings, improved performance through KDTree implementation replacing mdtraj.compute_contacts to mitigate memory errors, and updates to the DPM solver. The release also adds input sequence validity checking, changes the default random seed to use system time instead of zero, and includes a toy example of PPFT training along with various documentation improvements.
https://github.com/microsoft/bioemu/releases/tag/v.1.2.0
BoltzGen 0.1.8 fixes antibody-related documentation and file names while adding the "antibody-anything" protocol to the CLI. This release includes performance improvements with clamped analyze worker thread pools to prevent oversubscribing, SLURM example configurations, and training fixes along with model checkpoint loading updates.
https://github.com/HannesStark/boltzgen/releases/tag/v0.1.8
ChimeraX daily build 20251126 was added for Linux (the daily build for Mac was added last month), which allows users to run Botlz2 through ChimeraX natively.
https://www.rbvi.ucsf.edu/trac/ChimeraX/wiki/ChangeLog
DOCK 3.8 includes new mol2 output with full atom typing and bond information, improved sampling algorithms with better control, multiple ligand pose output, decreased memory usage, and improved hydroxyl sampling. The update also adds per-atom scoring function breakdowns and removes clashed conformation generation.
https://dock.compbio.ucsf.edu/DOCK3.8/index.html
Google Cloud SDK 548 adds the gcloud run compose up command, support for the --readiness-probe flag for gcloud run services command groups, --no-build flag to gcloud run deploy, and --source command in beta.
https://docs.cloud.google.com/sdk/docs/release-notes
HADDOCK3 2025.11.0 includes updates to dependencies and minor bug fixes.
https://github.com/haddocking/haddock3/releases/tag/2025.11.0
NAMD 3.0.2 point release fixes bugs in the Colvars module that impact versions 3.0, 3.0.1, and earlier NAMD 2 versions.
https://www.ks.uiuc.edu/Research/namd/3.0.2/announce.html
OpenFold 2.2.0 introduces multimer inference support using AlphaFold-Multimer weights, single-sequence inference (SoloSeq) using ESM-1b embeddings, and a custom DeepSpeed DS4Sci_EvoformerAttention kernel providing 13x reduced peak memory and 4x inference speedup. This release also includes upgrades to PyTorch 2, CUDA 12, NumPy 2, and PyTorch Lightning 2.5, adds comprehensive documentation at ReadTheDocs, ModelCIF output format support, and numerous training stability improvements including FP16 fixes.
https://github.com/aqlaboratory/openfold/releases/tag/v2.2.0
ORCA 6.1.1 irons out glitches that remained following the release of ORCA 6.1 and improves documentation.
https://www.faccts.de/customer
pytom-match-pick 0.12.1 includes bugfixes for argparse validation and corrects the conversion factor for microns to meters in Warp XML files.
https://github.com/SBC-Utrecht/pytom-match-pick/releases/tag/0.12.1
pyTME 0.3.3 introduces on-the-fly score background correction for commonly used scores and extends analysis support for pytme_runner to simplify large dataset analysis. The update includes extended analyzer support for the Jax backend with additional aggregation operations like constrained matching, along with internal improvements to filter composition and transformation matrix handling.
https://github.com/KosinskiLab/pyTME/releases/tag/v0.3.3
SamTOOLS 1.23 adds new features, including richer reference statistics in samtools stats, faster handling of reference caches, new depth filtering for samtools coverage, and better control over read groups, UMIs, and soft-clipped reads in FASTQ output. Performance and usability are improved with more efficient file splitting and clearer help and documentation, including new support information. This update also fixes some bugs, cleans up memory leaks, and introduces code quality improvements in general.
https://github.com/samtools/samtools/releases/tag/1.23
Schrödinger 2025-3 introduces performance improvements including nearly 2x faster Glide docking, 5x faster FFBuilder with the new MPNICE potential, and enhanced AI capabilities with the new Maestro Assistant beta for conversational help and natural language commands. The release adds comprehensive nucleic acid modeling workflows, an improved FEP+ settings interface, machine learning force field (MLFF) support across multiple modules, and expanded materials science capabilities including ionic conductivity predictions and coarse-grained molecular dynamics enhancements.
https://www.schrodinger.com/life-science/download/release-notes/release-2025-3/
Schrödinger 2025-4 includes enhancements across small molecule and biologics drug discovery, materials science, and platform capabilities. Key updates include interactive Core Hopping in Ligand Designer for exploring diverse chemical space, full release of TCR structure prediction through TCRBuilder2, expanded Mixed Solvent MD with immiscible probes for cryptic pocket identification, and new defect formation energy workflows in Quantum ESPRESSO. This release also introduces Form Templates in LiveDesign, improved Maestro Assistant functionality, and comprehensive coarse-grained modeling capabilities for complex bilayer systems.
https://www.schrodinger.com/life-science/download/release-notes/
Topaz 0.3.17 is now available. There is no changelog for this release.
https://github.com/tbepler/topaz/releases/tag/v0.3.17
