A Bayesian approach to extracting free-energy profiles from cryo-electron microscopy experiments

Abstract

Cryo-electron microscopy (cryo-EM) extracts single-particle density projections of individual biomolecules. Although cryo-EM is widely used for 3D reconstruction, due to its single-particle nature it has the potential to provide information about a biomolecule’s conformational variability and underlying free-energy landscape. However, treating cryo-EM as a single-molecule technique is challenging because of the low signal-to-noise ratio (SNR) in individual particles. In this work, we propose the cryo-BIFE method (cryo-EM Bayesian Inference of Free-Energy profiles), which uses a path collective variable to extract free-energy profiles and their uncertainties from cryo-EM images. We test the framework on several synthetic systems where the imaging parameters and conditions were controlled. We found that for realistic cryo-EM environments and relevant biomolecular systems, it is possible to recover the underlying free energy, with the pose accuracy and SNR as crucial determinants. We then use the method to study the conformational transitions of a calcium-activated channel with real cryo-EM particles. Interestingly, we recover not only the most probable conformation (used to generate a high-resolution reconstruction of the calcium-bound state) but also a metastable state that corresponds to the calcium-unbound conformation. As expected for turnover transitions within the same sample, the activation barriers are on the order of $$k_BT$$ k B T . We expect our tool for extracting free-energy profiles from cryo-EM images to enable more complete characterization of the thermodynamic ensemble of biomolecules.

Recovery of Conformational Continuum From Single-Particle Cryo-EM Images: Optimization of ManifoldEM Informed by Ground Truth

“…Specifically, an optimization approach could be designed to compare images within each bin and reassign erroneously-assigned snapshots into neighboring bins in which they most likely belong. To note, a maximum-likelihood approach does already exist that aims to extract such granular conformational heterogeneity [ 42 ], as does a method based on neural networks [ 32 ]. A more comprehensive discussion of additional, less-impactful improvements to the core ESPER method is also available [ 41 ].…”

“…We first introduce our framework for the creation of synthetic ground-truth single-particle cryo-EM data sets in the form of 2D projections of 3D density maps arising from a quasi-continuum of atomic structures [ 39 ], [ 40 ]. In the time since its conception, this synthetic frameworkhas already been used as a performance benchmark by two other groups [ 29 ], [ 42 ]. To begin, a suitable macromolecule is chosen as a foundational model, defined by structural information available in the form of 3D atomic coordinates from the Protein Data Bank (PDB) [ 43 ].…”

DeepHEMNMA: ResNet-based hybrid analysis of continuous conformational heterogeneity in cryo-EM single particle images

“…The last decade was marked by an active research in methods to pave the way for a full exploration of larger degrees of continuous conformational heterogeneity ( Dashti et al, 2014 ; Jin et al, 2014 ; Tagare et al, 2015 ; Haselbach et al, 2018 ; Dashti et al, 2020 ; Harastani et al, 2020 ; Lederman et al, 2020 ; Moscovich et al, 2020 ; Giraldo-Barreto et al, 2021 ; Punjani and Fleet, 2021 ). These methods aim at determining the full conformational distribution (also called conformational space, landscape, or manifold), based on which the images with similar conformations could be assembled in 3D reconstructions and, optionally, a displacement of a 3D model can be animated in this space without calculating 3D reconstructions ( Jin et al, 2014 ; Harastani et al, 2020 ).…”

Protein dynamics developments for the large scale and cryoEM: case study of ProDy 2.0

“…A trend towards customisable pipelines, such as SignDy and CryoDy, makes these techniques even more widely usable and we expect great developments in the future, aided by continuing developments in structural biology including the availability of structural models resolved by AlphaFold2 (Jumper et al, 2021;Varadi et al, 2021). The next big area is clearly continuous heterogeneity/dynamics analysis of cryoEM images (Chen & Ludtke, 2021;Giraldo-Barreto et al, 2021;Herreros et al, 2021;Sorzano et al, 2019) in place of the existing discrete classification approaches, which could benefit from a better connection to such computational biophysics approaches. 2018-SyG, Proposal 810057) and 'ERDF A way of making Europe' from the European Union and Horizon 2020, and PID2019-104757RB-I00 funded by MCIN/AEI to JMK, JMC and COSS and National Institutes of Health (NIH) grants R01 GM139297 and P41 GM103712 to IB.…”