SBGrid Newsletter: October 2025

Dear Consortium Members and Affiliates,

October SBGrid news includes links to watch our October Rocket webinar register for the November TomoNet webinar, a reminder about Coot issues on the new macOS, a Research Associate position posting at SBGrid, a software push with 11 updates and 11 new titles, and two publication highlights.

We had a great turnout for the first software webinar of the season. If you missed the live presentation on Rocket, you can watch the video on SBGrid's YouTube channel 

Next up: join us November 18th to hear from Hui Wang, postdoc in the Zhou Lab at University of California, Los Angeles, on TomoNet: Automatic Particle Picking of Flexible Lattice-Like Structures for CryoET.

To receive email reminders about upcoming webinars, please be sure to register for the series! Registration here: https://sbgrid.org/webinars/#register

November 18: Hui Wang - TomoNet: Automatic Particle Picking of Flexible Lattice-Like Structures for CryoET

December 9: to be announced

January 13: Alice-Roza Eruera - Building CryoET Training Workflows Using the SBGrid Ecosystem

February 10: Open Free Energy

March 10:  Open Force Field

April 14: Jose Luis Vilas - Image processing made it simple with scipionTomo

May 12: Jarrett Johnson - PyMOL

June 9: to be announced

Webinar registration and details

SBGrid webinars are hosted with partial support from the NIH R25 Continuing Education for Structural Biology Mentors #GM151273, in collaboration with Co-PI Jamaine Davis of Meharry Medical College.

This month's software push includes updates to ChimeraX, DeepRank-GNN-ESM, MDAnalysis, NGLView, NmrPipe, Open Force Field Toolkit, Open Free Energy,  PowerFit, PyTOM-match-pick, RDKit, and TomoNet along with these new titles: Ais, ANARCI, BoltzDesign, CTFMeasure, Icecream, nvMolKit, OpenEye Academic Package (for licensed academic labs), Pom, RoseTTAFold2, ROCKET, and SimpleFold. See Software Changes below for complete details.

Coot issues in MacOS Tahoe: A reminder to Coot users that we recommend waiting to upgrade to the latest macOS (Tahoe) due to problems with XQuartz-based OpenGL acceleration. We are monitoring the issue and will provide updates when available.

Research Associate - SBGrid, Harvard Medical School

We're looking for someone to join the SBGrid team as a Research Associate to provide support to our community of users - you! This position is for a computationally-minded PhD structural biologist with 2+ years hands-on experience doing CryoEM workflows/analysis, and 3+ years command-line experience in Unix-based operating systems (Linux, macOS). An added bonus is someone with programming ability (Python, shell), experience with software development and deployment tools, and familiarity with other structure determination or computational techniques. Please see the full description on the Harvard website: 

https://academicpositions.harvard.edu/postings/15185.

Over 100 new member publications appeared in journals this month. You can find a complete listing on our website, along with a couple of notable highlights below:

- Caltech doctoral student Vida Storm Robertson highlighted a publication from John Williams of City of Hope and colleagues, who developed a new coronavirus biologic that tightly binds to SARS-CoV-2 spike proteins. Read more: https://medium.com/sbgrid-community-news/out-competing-covid-3346b524994a

- Recent Ph.D. graduate KeAndreya Morrison (congratulations KeAndreya!) chose a publication from the laboratory of Walter Chazin to highlight. Chazin and colleagues uncover how a flexible tail of the immune protein calprotectin shapes zinc binding and bacterial growth, revealing new insights into our body’s fight against Staphylococcus aureus. Read more: https://medium.com/sbgrid-community-news/how-one-protein-tail-helps-starve-bacteria-of-zinc-bcaf646749a5

If you're currently preparing a manuscript, please remember that, while you're making the PDB record deposit and publication submission, you can also preserve your primary experimental datasets with deposits to the SBGrid Data Bank.

SBGrid operations are funded with member fees and grants, so we are grateful when you are able to acknowledge SBGrid in your presentations and publications.

Please use this SBGrid logo on the acknowledgements slide of your presentations.

We recommend the following boilerplate language for inclusion in publications that report results obtained with SBGrid supported software:

Structural biology applications used in this project were compiled and configured by SBGrid [1].

[1] A. Morin, B. Eisenbraun, J. Key, P. C. Sanschagrin, M. A. Timony, M. Ottaviano, and P. Sliz, “Collaboration gets the most out of software.,” Elife, vol. 2, p. e01456, Sep. 2013.

Link to article: https://elifesciences.org/articles/01456.

SBGrid's eLife paper received seven new citations since our last reporting, in publications with contributions from SBGrid member labs:

Halil Aydin from New York University and Matteo Dal Peraro from École Polytechnique Fédérale de Lausanne in Nature Communications: 

Cardiolipin dynamics promote membrane remodeling by mitochondrial OPA1;

Goran Bajic from Icahn School of Medicine at Mount Sinai in Science Immunology:

Germinal center–mediated broadening of B cell responses to SARS-CoV-2 booster immunization;

A review in Nature Reviews Methods Primers with contributions from many member labs and contributing developers, including Pavel Afonine from Lawrence Berkeley National Laboratory, Kay Diederichs from University of Konstanz, Eugene Krissinel from Rutherford Appleton Laboratory, José Antonio Márquez from EMBLGrenoble, Andrea Thorn from Hamburg University, and Isabel Usón from IBMB, CSIC:

Macromolecular crystallography;

Kevin Campbell from University of Iowa in Nature: 

LARGE1 processively polymerizes length-controlled matriglycan on prodystroglycan;

Jean-Philippe Julien from The Hospital for Sick Children Research Institute in Immunity:

Cryo-EM structure of endogenous Pfs230:Pfs48/45 complex with six antibodies reveals mechanisms of malaria transmission-blocking activity;

George Phillips Jr. in Journal of Natural Products:

Functional and Structural Studies on the Esperamicin Thioesterase and Progress toward Understanding Enediyne Core Biosynthesis; and

Sichen Shao from Harvard Medical School in Science: 

GCN1 couples GCN2 to ribosomal state to initiate amino acid response pathway signaling.

Ais 1.0.44 is a fast and user-friendly annotation and segmentation tool for cryoET data using convolutional neural networks. The software enables concurrent annotation of multiple features (membranes, ribosomes, mitochondrial granules, microtubuli) and includes model training capabilities with interaction filtering to reduce false positives. It provides built-in 3D rendering for result inspection and supports batch processing for automated segmentation of multiple tomograms.

https://github.com/bionanopatterning/Ais/tree/master

ANARCI 2024.05.21 is a tool for Antibody Numbering and Antigen Receptor ClassIfication for a variety of species.

https://github.com/oxpig/ANARCI

BoltzDesign 20250714 is a molecular design tool powered by the Boltz model for designing protein-protein interactions and biomolecular complexes. It features an end-to-end pipeline that includes sequence optimization using ProteinMPNN/LigandMPNN, structure validation with AlphaFold3, and support for small molecules, DNA, and RNA targets. The tool provides trajectory visualization capabilities and automated setup for complete molecular design workflows.

https://github.com/yehlincho/BoltzDesign1

CTFMeasure 1.4.0 is a software for estimating CTF parameters of cryo-ET tilt series. It features tilt-series-based joint estimation and Thon ring rescaling technique to retrieve power spectrums with higher SNR and more clear Thon rings, leading to higher estimation accuracy. For focused ion beam (FIB) milling lamellas, the absolute tilt angle offset of the lamellar sample relative to the sample stage plane can be estimated together with the CTF parameters.

https://thuem.net/software/ctfmeasure/releaseNote.html 

Icecream 2644dc2 is a self-supervised framework for cryo-ET reconstruction that integrates equivariance principles from modern imaging theory into a deep-learning architecture. Icecream provides a theoretically grounded and computationally efficient method that jointly performs denoising and missing-wedge correction. The current version supports training on a single split of the tomograms. Upcoming updates will include support for multi-volume training.

https://github.com/swing-research/icecream?tab=readme-ov-file#-note 

nvMolKit 0.2.0 is a CUDA-backed Python library for accelerating common RDKit molecular operations through batch-oriented GPU processing. The library provides GPU-accelerated Morgan fingerprints, molecular similarity calculations, ETKDG conformer generation, and MMFF geometry relaxation while maintaining compatibility with RDKit's API. It returns asynchronous GPU results that can be converted to PyTorch tensors or NumPy arrays, and supports multi-GPU processing for computationally intensive operations.

https://nvidia-digital-bio.github.io/nvMolKit/

OpenEye Academic Package 2025.1.1 is available to nonprofit groups with a license. This collection of tools helps with design of potent and selective bioactive molecules. See instructions at https://sbgrid.org/software/titles/openeye to include this package in your collection.

https://docs.eyesopen.com/applications/notes/releasenotes/highlights2025_Spr.html  

Pom a804c9d uses comprehensive segmentation to scale data analyses in cellular cryoET.

https://github.com/bionanopatterning/Pom 

RoseTTAFold2 1.0 extends the original three-track architecture of RoseTTAFold over the full network, incorporating the concepts of Frame-aligned point error, recycling during training, and the use of a distillation set from AlphaFold2.With the addition of RoseTTAFold2 the pipeline for RFantibody, for the structure-based design of de novo antibodies and nanobodies, is complete in SBGrid. 

https://github.com/uw-ipd/RoseTTAFold2/releases/tag/v1.0 

https://github.com/RosettaCommons/RFantibody 

ROCKET 0.0.1 is a tool for refining AlphaFold predictions using crystallographic and cryo-EM likelihood targets. It integrates with OpenFold to condition experimental structure determination on pretrained neural networks, requiring both OpenFold and Phenix installations for automatic data preprocessing and post-refinement geometry polishing. The software provides a framework for improving protein structure predictions by incorporating experimental data constraints.

https://github.com/alisiafadini/ROCKET

SimpleFold 0.1.0 is a flow-matching-based protein folding model that uses only general purpose transformer layers. The model scales to 3B parameters and is trained on over 8.6M distilled protein structures plus experimental PDB data, making it the largest scale folding model ever developed. SimpleFold achieves competitive performance on standard folding benchmarks and demonstrates strong ensemble prediction capabilities due to its generative training objective.

https://github.com/apple/ml-simplefold

ChimeraX 20251016-arm64 is a daily build and allows users to run Botlz2 through ChimeraX natively on Macs with an M processor.

https://www.rbvi.ucsf.edu/trac/ChimeraX/wiki/ChangeLog 

DeepRank-GNN-ESM 0.2.3 fixes output file path generation in hdf5_to_csv.py and adds support for building both CPU and GPU Docker images. This release addresses file handling issues in the data conversion pipeline.

https://github.com/haddocking/deeprank-gnn-esm/releases/tag/v0.2.3

MDAnalysis 2.10.0 adds support for NumPy 2.0+ while maintaining backwards compatibility to NumPy 1.26.0, and introduces new parallelization support for LinearDensity, PersistenceLength, and InterRDF analyses. This release includes enhanced trajectory support with dt reading for XTC/TRR files, improved GROMOS11 reader performance, and new capabilities for interactive MD stream reading and non-linear time averaged MSD calculations.

https://github.com/MDAnalysis/mdanalysis/releases/tag/package-2.10.0

NGLView 4.0 includes major upgrades to molstar 4.9.0 and NGL 2.4.0, along with fixes for pkg_resources deprecation warnings, rotation functionality, and trajectory embedding warnings. The release also adds Python 3.13 support and refactors widget code for improved maintainability.

https://github.com/nglviewer/nglview/releases/tag/v4.0

NmrPipe 20250729 is a new update of this title for processing, analyzing, and exploiting multidimensional NMR spectroscopic data (no release notes/changelog).

Open Force Field Toolkit 0.17.1 is a minor feature and bugfix release that addresses various issues and improvements in the molecular simulation toolkit for force field development and application.

https://github.com/openforcefield/openff-toolkit/releases/tag/0.17.1

Open Free Energy 1.6.1 includes improvements to partial charge tests with production NAGL model support, debugging of slow Python 3.13 tests, and enhanced documentation with quickrun execution updates and jq cookbook examples. The release also addresses build issues and removes unnecessary residue ID limitations.

https://github.com/OpenFreeEnergy/openfe/releases/tag/v1.6.1

OpenFreeEnergy 1.7.0 introduces the new SepTop Protocol for separated topology free energy calculations, implements single residue Boresch host finding with improved restraint settings, and adds the ABFE (Absolute Binding Free Energy) protocol. This release also migrates to Pydantic v2, makes CUDA the default compute platform, and includes numerous improvements to host-guest restraint handling and general performance optimizations.

https://github.com/OpenFreeEnergy/openfe/releases/tag/v1.7.0

PowerFit 4.0.1 fixes a sigma difference calculation bug that was affecting rigid body fitting accuracy. This minor release maintains the tool's core functionality for fitting atomic structures like proteins into cryo-electron microscopy density maps.

https://github.com/haddocking/powerfit/releases/tag/v4.0.1

PowerFit 4.0.2 includes minor bug fixes.

PyTOM-match-pick 0.11.1 fixes symlink extraction by ensuring absolute paths no longer resolve symlinks or normalize paths, addressing path handling issues in the template matching and particle picking workflow.

https://github.com/SBC-Utrecht/pytom-match-pick/releases/tag/0.11.1

PyTOM-match-pick 0.12.0 includes API-breaking changes that require tilt angles for TMJob initialization and introduces a new tilt-series dataclass for better data organization. This release fixes critical bugs in tomogram wedge computation for volume split mode and eliminates incorrect wedge calculations at 90-degree tilt angles.

https://github.com/SBC-Utrecht/pytom-match-pick/releases/tag/0.12.0

RDKit 2025.09.1 introduces SCSR (self-contained sequence representation) file format parsing, updated CDXML parser with new CDX file support, and multiple shape-based alignment improvements. This release switches to C++20, includes backwards incompatible changes to atom map handling and 2D coordinate generation, and provides numerous bug fixes including SMILES canonicalization issues and multithreaded reader improvements.

https://github.com/rdkit/rdkit/releases/tag/Release_2025_09_1

TomoNet 1.04 fixes known issues fixed and support for Relion5 format added.

https://github.com/logicvay2010/TomoNet/releases 

Please note that not all software applications are available to every SBGrid member type. If you see an application that you would like to use, but is not included in your software tree, please contact us to find out what options are available for access.

This newsletter is sent to you because you are a member or affiliate of the SBGrid Consortium.

More information about the SBGrid Consortium is available at https://sbgrid.org

Report software bugs: sbgrid.org/bugs