Mask-based detection method for three-dimensional density map resolution of cryo-electron microscopy

Abstract

The invention discloses a method for detecting resolution of a three-dimensional density picture of a cryo-electron microscopy based on a mask. The method is characterized in that the three-dimensional density picture reconstructed by the cryo-electron microscopy is segmented through a three-dimensional binary mask, radius power spectrums of an intramembranous particle structure part and an extramembranous background noise part of the three-dimensional density picture based on a three-dimensional shell of a Fourier space are respectively computed to obtain signal power and noise power of the three-dimensional density picture, so that an estimated value of the global resolution is obtained through a spectral signal to noise ratio based on the mask, which is obtained through calculation of the spectral signal to noise ratio, and a selected threshold, and thus evaluation of the reconstruction quality of the cryo-electron microscopy is implemented. According to the method, the resolution can be directly estimated from the three-dimensional density picture reconstructed by the cryo-electron microscopy, an algorithm is simple and practicable, and the method has a larger application space.

Description

technical field [0001] The invention relates to a technology in the field of image processing, in particular to a method for detecting the resolution of a cryo-electron microscope three-dimensional density map based on a mask. Background technique [0002] Cryo-electron microscopy (Cryo-EM) technology has become an essential research tool in structural biology. Compared with traditional X-ray (X-ray) and nuclear magnetic resonance (NMR), cryo-electron microscopy has specific advantages, such as: (1) It does not need to grow crystals, so the molecular structure solved is closer to its natural physiological state; (2) ) can be used to solve the three-dimensional structure of complexes with huge molecular weight, viruses, and even organelles; (3) it can study the dynamic structure of active biological macromolecules. In recent years, with the development of electron microscopes, data acquisition cameras, efficient algorithms and programs, and high-performance computers, cryo-e...

AI Technical Summary

Problems solved by technology

However, this method has the following defects: the method relies on the reconstruction process, and is susceptible to noise and overfitting; the structural quality of the two independent half-dataset reconstructions is lower than that of the full data set reconstruction, which is a redundant calculation. process; estimating resolution from FSC curves requires a threshold standard, and there is no unified threshold standard at present

However, this technology needs to obtain the average value and variance of the noise first, and then obtain the average value and variance of the signal based on this. The calculation error of the noise is brought into the calculation of the signal, and iterative fitting is required. The process is complicated and there is a risk of error

Images