RELION 3.1.0 Now Available

RELION 3.1.0 was released on July 3, 2020. Everyone using RELION 3.0.x or 3.1 beta should upgrade to this version.

You can download the newest release here.

Major changes since RELION 3.0.x include:

  • High-order aberration correction & anisotropic magnification correction in CTF refinement
  • External jobtype to run third-party software in the RELION pipeline
  • Schedules for designing & executing on-the-fly processing procedures
  • Tweaks to GUI, pipeliner and helical processing
  • Symmetry relaxation (special thanks to Vahid Abrishami for code contribution)

The new tutorial is available at

Support for the EER movie format is not yet included in this release. Please use a special version in the devel-eer branch at This feature will be merged in near future as a minor update (version 3.1.x).

What is RELION?

RELION (for REgularised LIkelihood OptimisatioN, pronounce rely-on) is a stand-alone, open source computer program that employs an empirical Bayesian approach to refinement of (multiple) 3D reconstructions or 2D class averages in electron cryo-microscopy (cryo-EM). It is developed in the group of Sjors Scheres at the MRC Laboratory of Molecular Biology. Whereas alternative approaches often rely on user expertise for the tuning of parameters, RELION uses a Bayesian approach to infer parameters of a statistical model from the data.

Interested in getting faster cryo-EM results?
Learn more about RELION GPU Accelerated Systems for Cryo-EM

Options for accelerated versions

Parts of the cryo-EM processing pipeline can be very computationally demanding, and in some cases special hardware can be used to make these faster. There are two such cases at the moment;

  • GPU acceleration: RELION only supports CUDA-capable GPUs of compute capability 3.5 or higher.
  • Vectorized CPU code path: RELION only supports GCC and ICC 2018.3 or later.

Note that you cannot have both acceleration in the same binary at the moment.

There are more benefits than speed; the accelerated versions also have a decreased memory footprint. Details about how to enable either of these options is listed below.


Tools that are GPU-accelerated:

  • relion_refine (i.e. Class2D, Class3D, Refine3D, Multibody refinement)
  • relion_autopick

Classification without alignment is not accelerated.

When CUDA SDK is available, GPU support is automatically compiled.


If you run relion_refine with a the “--gpu” flag, you will run the accelerated CUDA version of the kernels. If you leave out the “--gpu” flag, it will run the original CPU version.


Tools that are CPU-accelerated (vectorized):

  • relion_refine (i.e. Class2D, Class3D, Refine3D, Multibody refinement)

Classification without alignment is not accelerated.

To build with support for CPU-accelerated kernels in addition to the original CPU version, build by setting ALTCPU=ON

cd build
rm -r *
cmake -DALTCPU=ON ..
make install

This will require the Intel TBB (Threading Building Blocks) library. RELION will look for TBB, and fetch and install it when it is missing on your system. You can force this behavior (and make sure you are using the latest version) by adding:


In addition, you can make use the Intel Math Kernel Library (Intel MKL). This is optional (but will scale better with increased threads). Add this by:



If you run relion_refine with a the “--cpu” flag, you will run the accelerated version. If you leave it the original CPU version will be run. You should use this flag if you can, unless you want to verify old runs or behavior.

For details on how to compile with Intel compilers and optimal runtime configurations, please look at their wiki.

Have any questions about RELION or other applications for molecular dynamics?
Contact Exxact Today