Our impact

We are dedicated to advancing the field of cryo-EM structure determination. To help others in their structure determination efforts we make all our algorithm developments available through our open-source program RELION, and provide access to video and slide materials of our locally organised EM-course. The journal Nature recognised our impact by listing Sjors as one of the ten people who mattered in 2014.

RELION

The empirical Bayesian approach to single-particle analysis that we developed has been implemented in an open-source program called RELION (for REgularised LIkelihood OptimisatioN). RELION may be downloaded for free from the RELION Wiki.

RELION played a crucial role in our demonstration that cryo-EM structures to near-atomic resolution may be obtained from only several tens of thousands of particles, and continues to produce great structures in our lab: see our structures for an overview. Many groups world wide have reported structures that were calculated using RELION. An interesting, early result came from the group of Joachim Frank, the original developer of the SPIDER software. Reprocessing of old, film data on a ribosomal pre-termination complex in his group led to a resolution improvement from 18 Angstroms in SPIDER to 9.7 Angstroms in RELION. The almost doubling of resolution compared to previously existing software illustrates that, apart from the major contribution of direct-electron detectors, RELION also played an important role in the cryo-EM "resolution revolution".

LMB EM-course

With the recent advances in cryo-EM structure determination, we find that an ever increasing number of people want to learn about it. To aid those new to the field, together with Lori Passmore and Paula daFonseca, we organized a course at LMB that covers the theoretical basis of modern cryo-EM structure determination. The slides and professionally edited videos of all ten lectures in this course may be downloaded for free from here.

News

RELION-2.0-beta released

Download from here

RELION provides a Bayesian approach to single-particle refinement. As such, it may be used to obtain reference-free 2D class averages, unsupervised 3D classifications, or state-of-the-art high-resolution refinements, all with minimal user intervention. The 2.0 release uses GPU acceleration to greatly enhance speed, implements a workflow engine, and has functionality for helical processing.