Access SBGrid Software

First, Check How You Access SBGrid Software

All SBGrid Consortium members have access to the SBGrid collection of structural biology applications on their own computers in one of two ways.

Shared/Multi-User Installation

Laboratory or institutional system administrators manage software installation and updates, frequently using network storage to provide SBGrid programs to more than one computer.

Personal Installer

Use the SBGrid Installation Manager to install software on your own workstation, laptop or container. Instructions: SBGrid Downloads.

🔹 Not sure which approach you're using, or which to setup makes the most sense for you? Email us at bugs@sbgrid.org and our team will guide you.

🔹 To confirm if your lab is an SBGrid member please check here: SBGrid Member List.

Operating System: macOS or Linux

Support for new releases of operating systems continues to evolve, and feedback from the user community is critical to ensure that we meet your needs. All applications are curated and tested on a limited number of core platforms, which receive full support and enable our software curators to efficiently troubleshoot any issues.

Some users have reported success running SBGrid applications on additional platforms, but this should be done with caution and at your own risk.

macOS

Generally simpler to administer than Linux; however Apple's approach to depreciation limits support for older software.

Linux

More complex to administer than macOS, but required for the majority of GPU-accelerated programs (CryoEM, CryoET, Structural Biology AI). We support RHEL, Rocky, and AlmaLinux platforms (other platforms, e.g. Ubuntu LTS/debian platforms might work, but are not tested by SBGrid).

Windows

SBGrid does not support Windows, but some users have had success using SBGrid with Windows Subsystem for Linux (WSL2).

Hardware Recommendations for SBGrid

Computer hardware to support SBGrid is typically selected in consultation with local system administrators and hardware vendors. Systems used for computational work generally require tiered storage (remote network project space combined with local SSD scratch space), NVIDIA GPUs, 64-128+ GB of memory, multicore processors, and 1-10 GbE networking. Less powerful laptops or desktops can be used effectively for routine structural analyses.

Many research groups have successfully deployed SBGrid on campus high-performance computing (HPC) resources, and an increasing number of structural biologists now run the SBGrid software collection in AWS cloud environments.

Most importantly, users should maintain a comprehensive data security and backup strategy—including redundant RAID arrays, data replication between systems, and offline backups. Data supporting published papers should be deposited in community repositories to ensure long-term accessibility and reproducibility.

For questions about system configuration or setup, please contact bugs@sbgrid.org.

Note that hardware configurations vary widely among laboratories and evolve regularly.