🚀Introducing fully automated data processing for repeat-target #cryoEM. Using new tools in #CryoSPARC, it is now possible to obtain resolutions & map quality equal to or better than manual processing, with zero user intervention. Preprint: biorxiv.org/content/10.1101/…

The automation strategy we have developed can be set up once for any specific target class (e.g. active state GPCRs with Nb35) and reused to process multiple datasets fully hands-off. It makes use of minimal prior knowledge through a low-res (15Å) 3D reference map.

We test across 21 challenging active and inactive GPCR datasets. Despite small particle size and the presence of contaminants, empty micelles, broken, denatured and aggregated particles, our automated workflow yields equal or better resolution and map quality in 17 of 21 cases.

In 10 of 21 cases, the automated results enable improved model building of the ligand, binding pocket, side chain positioning and/or extracellular loops. Three striking examples are shown here and in the first video.

The automated workflow generalizes easily to new target classes. We provide downloadable #CryoSPARC Workflow JSON files so that users can replicate, adapt and extend it for their own targets. All required jobs and tools are already available in v4.7.1+.

The workflow is powered by new methods, including micrograph junk detection, denoising, and reference based auto selection of 2D classes and 3D volumes. Read the preprint for more insights and details.