scholarly journals Deep Consensus, a deep learning-based approach for particle pruning in cryo-electron microscopy

IUCrJ ◽  
2018 ◽  
Vol 5 (6) ◽  
pp. 854-865 ◽  
Author(s):  
Ruben Sanchez-Garcia ◽  
Joan Segura ◽  
David Maluenda ◽  
Jose Maria Carazo ◽  
Carlos Oscar S. Sorzano

Single-particle cryo-electron microscopy (cryo-EM) has recently become a mainstream technique for the structural determination of macromolecules. Typical cryo-EM workflows collect hundreds of thousands of single-particle projections from thousands of micrographs using particle-picking algorithms. However, the number of false positives selected by these algorithms is large, so that a number of different `cleaning steps' are necessary to decrease the false-positive ratio. Most commonly employed techniques for the pruning of false-positive particles are time-consuming and require user intervention. In order to overcome these limitations, a deep learning-based algorithm named Deep Consensus is presented in this work. Deep Consensus works by computing a smart consensus over the output of different particle-picking algorithms, resulting in a set of particles with a lower false-positive ratio than the initial set obtained by the pickers. Deep Consensus is based on a deep convolutional neural network that is trained on a semi-automatically generated data set. The performance of Deep Consensus has been assessed on two well known experimental data sets, virtually eliminating user intervention for pruning, and enhances the reproducibility and objectivity of the whole process while achieving precision and recall figures above 90%.

IUCrJ ◽  
2020 ◽  
Vol 7 (6) ◽  
pp. 1179-1187 ◽  
Author(s):  
Jennifer N. Cash ◽  
Sarah Kearns ◽  
Yilai Li ◽  
Michael A. Cianfrocco

Recent advances in single-particle cryo-electron microscopy (cryo-EM) data collection utilize beam-image shift to improve throughput. Despite implementation on 300 keV cryo-EM instruments, it remains unknown how well beam-image-shift data collection affects data quality on 200 keV instruments and the extent to which aberrations can be computationally corrected. To test this, a cryo-EM data set for aldolase was collected at 200 keV using beam-image shift and analyzed. This analysis shows that the instrument beam tilt and particle motion initially limited the resolution to 4.9 Å. After particle polishing and iterative rounds of aberration correction in RELION, a 2.8 Å resolution structure could be obtained. This analysis demonstrates that software correction of microscope aberrations can provide a significant improvement in resolution at 200 keV.


2021 ◽  
Vol 4 (1) ◽  
Author(s):  
Blesson George ◽  
Anshul Assaiya ◽  
Robin J. Roy ◽  
Ajit Kembhavi ◽  
Radha Chauhan ◽  
...  

AbstractParticle identification and selection, which is a prerequisite for high-resolution structure determination of biological macromolecules via single-particle cryo-electron microscopy poses a major bottleneck for automating the steps of structure determination. Here, we present a generalized deep learning tool, CASSPER, for the automated detection and isolation of protein particles in transmission microscope images. This deep learning tool uses Semantic Segmentation and a collection of visually prepared training samples to capture the differences in the transmission intensities of protein, ice, carbon, and other impurities found in the micrograph. CASSPER is a semantic segmentation based method that does pixel-level classification and completely eliminates the need for manual particle picking. Integration of Contrast Limited Adaptive Histogram Equalization (CLAHE) in CASSPER enables high-fidelity particle detection in micrographs with variable ice thickness and contrast. A generalized CASSPER model works with high efficiency on unseen datasets and can potentially pick particles on-the-fly, enabling data processing automation.


2020 ◽  
Author(s):  
Blesson George ◽  
Anshul Assaiya ◽  
Robin Jacob Roy ◽  
Ajit Kembhavi ◽  
Radha Chauhan ◽  
...  

AbstractSingle-particle cryo-electron microscopy has emerged as the method of choice for structure determination of proteins and protein complexes. However, particle identification and selection which is a prerequisite for achieving high-resolution still poses a major bottleneck for automating the steps of structure determination. Here, we present a generalised deep learning tool, CASSPER, for the automated detection and isolation of protein particles in transmission microscope images. This deep learning tool uses Semantic Segmentation and a collection of visually prepared training samples to capture the differences in the transmission intensities of protein, ice, carbon and other impurities found in the micrograph. CASSPER is the first method to do pixel level classification and completely eliminates the need of manual particle picking. Integration of Contrast Limited Adaptive Histogram Equalization (CLAHE) in CASSPER enables high-fidelity particle detection even in micrographs with variable ice thickness and contrast. In addition, our generalized model for cross molecule picking works with high efficiency on unseen datasets and can potentially pick particles on-the-fly, thereby, enabling automation of data processing.


Science ◽  
2018 ◽  
Vol 361 (6405) ◽  
pp. 876-880 ◽  
Author(s):  
Yifan Cheng

Cryo–electron microscopy, or simply cryo-EM, refers mainly to three very different yet closely related techniques: electron crystallography, single-particle cryo-EM, and electron cryotomography. In the past few years, single-particle cryo-EM in particular has triggered a revolution in structural biology and has become a newly dominant discipline. This Review examines the fascinating story of its start and evolution over the past 40-plus years, delves into how and why the recent technological advances have been so groundbreaking, and briefly considers where the technique may be headed in the future.


2020 ◽  
Author(s):  
Jing Cheng ◽  
Bufan Li ◽  
Long Si ◽  
Xinzheng Zhang

AbstractCryo-electron microscopy (cryo-EM) tomography is a powerful tool for in situ structure determination. However, this method requires the acquisition of tilt series, and its time consuming throughput of acquiring tilt series severely slows determination of in situ structures. By treating the electron densities of non-target protein as non-Gaussian distributed noise, we developed a new target function that greatly improves the efficiency of the recognition of the target protein in a single cryo-EM image without acquiring tilt series. Moreover, we developed a sorting function that effectively eliminates the false positive detection, which not only improves the resolution during the subsequent structure refinement procedure but also allows using homolog proteins as models to recognize the target protein. Together, we developed an in situ single particle analysis (isSPA) method. Our isSPA method was successfully applied to solve structures of glycoproteins on the surface of a non-icosahedral virus and Rubisco inside the carboxysome. The cryo-EM data from both samples were collected within 24 hours, thus allowing fast and simple structural determination in situ.


2017 ◽  
Vol 8 (1) ◽  
Author(s):  
Linda E. Franken ◽  
Gert T. Oostergetel ◽  
Tjaard Pijning ◽  
Pranav Puri ◽  
Valentina Arkhipova ◽  
...  

2020 ◽  
Vol 12 (2) ◽  
pp. 349-354 ◽  
Author(s):  
Yuichi Yokoyama ◽  
Tohru Terada ◽  
Kentaro Shimizu ◽  
Kouki Nishikawa ◽  
Daisuke Kozai ◽  
...  

Sign in / Sign up

Export Citation Format

Share Document