Cryo-EM requires the protein sample to be frozen before it can be imaged using an electron microscope. This innovative technology and its application to structural biology is the focus of the 2017 Nobel Prize in Chemistry, contributing to a wave of new structural information on proteins that are difficult or impossible to prepare for X-ray crystallography. .. With the proliferation of cryo-EM, new methods and tools have emerged to improve technology and increase efficiency. Many people are currently working on how to prepare samples, which makes it easy to obtain high resolution structural information from each experiment. Once the sample is prepared, there are many automated tools developed to select the frozen particles to be processed. I got my PhD when I was investigating these tools. Student Mateusz Olek and his supervisor Peijun Zhang discovered an unusual problem with the help of Yuriy Chaban and Donovan Webb.Their findings were published in the journal structure..
Blind spot of automation
The automatic particle picker is designed to analyze images and automatically select the best particles for your experiment. While investigating the performance of these tools, the team noticed that some of the images had no selection. These voids could not be easily explained, especially if the team visually inspected the image and clearly confirmed the presence of particles. For some reason, the automatic particle picker wasn’t aware of a particular area of the image. This issue can affect cryo-EM experiments. If the viable particles are left behind by automated software, scientists may not be able to collect all the data needed for the experiment.
His team quickly noticed that the background of the image was inconsistent. There were dark and bright areas. This could explain some of the failures of automated software that rely on measuring the contrast of protein particles against the background. To perform the cryo-EM, the protein particles are suspended in a thin film of ice, so the team concluded that the background mismatch in the image was related to the difference in ice thickness. This has caused some problems for researchers using Cryo EM. First, one of the obvious solutions is to make the ice film more uniform. Unfortunately, great efforts have been made to improve the preparation of cryo-EM samples, but it is still very difficult to produce a uniform ice film. Ice thickness gradients are often seen in cryo-EM samples.
Start from scratch
Mateusz and the team started from scratch to develop new ways to combat the ice problem. They started by segmenting the various images and analyzing the background. This allowed the picker to identify the particles regardless of the background from the ice. This innovation meant that in a given experiment, researchers could reliably collect more structural information by analyzing more protein particles. But this was not the end of the road. Collecting more particles is of great value to researchers using cryo-EM, but the amount of ice affects the quality of the particles, thus affecting the quality of the reconstructable cryo-EM map. Give. Mateusz and his team knew that software could be further evolved by providing researchers with immediate information about the quality of imaged protein particles.
In a recent publication, the research team highlighted two major issues that can occur when ice is not optimally thick. First, very thick ice increases the background noise of the image. This makes it difficult to obtain high resolution data from protein particles. Conversely, if the ice film is too thin, it may not be able to properly support the protein. This problem is highly dependent on the particular protein the researcher is investigating. For example, some proteins are small and tightly bound in a spherical structure. These proteins can be supported by a relatively thin layer of ice. However, many proteins are large, have long protruding branches, and if they are too thin, they can fall outside the ice film. This means that even with very low background noise, thin ice makes it difficult to image these types of proteins. Optimal ice thickness minimizes background, but is thick enough to fully support the protein. In many cases, the optimum ice thickness is protein It is under research and will change depending on the experiment.
Their new software tool, IceBreaker, is believed to be able to fully automate particle selection. However, the team chose an approach that would give more flexibility to the researchers who use it. Instead of making a decision, the software annotates each particle to show the researcher the quality. This allows you to proceed with the experiment exactly as you need it to achieve the best results. Less high-quality particles may be sufficient, or lower-quality particles from thicker ice may be required to exhibit a unique particle orientation that is not supported by thin ice. With the new IceBreaker software, researchers have complete control over the data they are collecting. IcreBreaker is currently implemented as part of Diamond’s data collection pipeline. Electron Bioimaging Center (eBIC) It’s available for free on GitHub.
Mateusz Olek et al, IceBreaker: Software for high resolution single particle cryo EM with non-uniform ice, structure (2022). DOI: 10.1016 / j.str.2022.01.005
Diamond light source
Quote: Https: //phys.org/news/2022-02-ice-problem-cryo-electron-microscopy.html by cryo-electron microscopy (February 14, 2022) obtained on February 14, 2022. Solving the ice problem
This document is subject to copyright. No part may be reproduced without written permission, except for fair transactions for personal investigation or research purposes. Content is provided for informational purposes only.
Solving the Ice Problem in Cryogenic Electron Microscopy
Source link Solving the Ice Problem in Cryogenic Electron Microscopy