Streamlining in-situ structure determination using cryo-ET

An open-source framework for analysis and visualization of cryo-electron tomography data to determine protein structures imaged within the native context of cells at molecular resolution.
Streamlining in-situ structure determination using cryo-ET

The ability to map the interior of cells at high spatial resolution to uncover the inner workings of life is both an exciting and challenging feat. Cryo-electron tomography (ET) has emerged as the technique of choice to determine the structure of proteins in their native, frozen hydrated state, where repetitive or frequently occurring molecules can be averaged in 3D to create detailed structural maps. Indeed, this technology has been successfully applied to study sub-cellular complexes present within intact bacteria or eukaryotic cells, paving the way for in-situ structural studies to soon become routine.

Challenges with molecular crowdedness and big data

Recent advances in automation during specimen preparation and strategies for high-throughput acquisition of tilt-series can already produce extraordinary amounts of data [1]. Analyzing this  data, however, requires access to prohibitive amounts of storage and compute resources, posing major barriers to the development and broader adoption of this technology [2]. In addition, seemingly simple tasks such as searching for tiny molecules within cells are challenging to execute due to the molecular crowding and the noise present in cryo-ET data.

Cryo-ET data analysis pipeline. Multiple steps are required to convert the raw tilt-series data into high-resolution structures by sub-tomogram averaging, a process that can take weeks to months to complete.

Democratizing access to high-resolution cryo-ET

To address these problems, we developed nextPYP, an open-source platform that provides a user-friendly interface to execute advanced cryo-ET workflows including real-time preprocessing, high-resolution refinement and conformational heterogeneity analysis. The package features a lightweight architecture with small storage footprint that in combination with high-performance distributed processing provides a scalable solution capable of streamlining the analysis and visualization of large in-situ datasets. Using a combination of established image analysis routines and new machine learning algorithms, nextPYP offers a comprehensive end-to-end solution for cryo-ET data processing, effectively broadening access to this powerful technology.

Visual proteomics, a step closer

To routinely generate molecular atlases of cells by providing the position and angular orientation of protein complexes with exquisite level of detail constitutes an exciting frontier in cellular biology. For example, studying translating states of ribosomes in the cellular context may provide valuable insights into protein biogenesis that could improve our understanding of biology and help combat disease. Ultimately, ubiquitous access to large datasets combined with robust and accessible algorithms will help realize the full potential of visual proteomics.


[1] Bouvette et al., Beam image-shift accelerated data acquisition for near-atomic resolution single-particle cryo-electron tomography. Nature Communications 12, 1957 (2021).

[2] Liu et al., High-resolution structure determination using high-throughput electron cryo-tomography. Acta Crystallographica Section D: Structural Biology 78 (7), 817-824 (2022).

For more information, visit

Please sign in or register for FREE

If you are a registered user on Research Communities by Springer Nature, please sign in