Macromolecule Particle Picking and Segmentation of a KLH Database by Unsupervised Cryo-EM Image Processing.

Miguel Carrasco,Nicole D Tischler,Patricio Toledo

doi:10.3390/biom9120809

Miguel Carrasco, Nicole D Tischler + Show 1 more

Open Access

https://doi.org/10.3390/biom9120809

Copy DOI

Abstract

Segmentation is one of the most important stages in the 3D reconstruction of macromolecule structures in cryo-electron microscopy. Due to the variability of macromolecules and the low signal-to-noise ratio of the structures present, there is no generally satisfactory solution to this process. This work proposes a new unsupervised particle picking and segmentation algorithm based on the composition of two well-known image filters: Anisotropic (Perona–Malik) diffusion and non-negative matrix factorization. This study focused on keyhole limpet hemocyanin (KLH) macromolecules which offer both a top view and a side view. Our proposal was able to detect both types of views and separate them automatically. In our experiments, we used 30 images from the KLH dataset of 680 positive classified regions. The true positive rate was 95.1% for top views and 77.8% for side views. The false negative rate was 14.3%. Although the false positive rate was high at 21.8%, it can be lowered with a supervised classification technique.

Highlights

Macromolecules consisting of proteins and nucleic acid play a crucial role in all living systems, and information on their structures is essential for achieving detailed mechanistic insights into their function
We briefly introduce the necessary concepts associated with noise-reduction techniques and the particular characteristics of cryo-electron microscopy (cryo-EM) images before explaining in detail the method proposed and analyzing the experimental tests
The Perona–Malik (PM) technique allows for noise level reductions and, at the same time, border preservation from structures through anisotropic diffusion constant tuning [23,30] based on solutions of the heat equation, meaning that the diffusion constant is lower near the border and higher in uniform areas

Summary

Introduction

Macromolecules consisting of proteins and nucleic acid play a crucial role in all living systems, and information on their structures is essential for achieving detailed mechanistic insights into their function. Atomic level high-resolution structures can reveal antigenic surfaces and molecular interaction sites such as those involved in multimerization and binding to substrates or other molecules. Structure determination by cryo-electron microscopy (cryo-EM) linked to 3D image reconstruction has reached near-atomic resolution thanks to Bayesian image processing algorithms and recent technological advances such as direct electron detectors [1,2,3,4]. High-resolution structure determination by cryo-EM demands the processing of thousands of single-particle images and, picking single particles from electron micrographs is still considered a difficult problem and most of the time is performed manually [1,5,6]. Micrographs often suffer from image distortions introduced by the microscope or detection systems and, may include heterogeneous particles which generate different 2D views in random orientations requiring classification [6,7,8]

Objectives

Methods

Results

Discussion

Conclusion