2025/02/05 – Roberto Covino

Roberto Covino
Professor of Computational Life Science, Goethe University Frankfurt
Senior Fellow, Frankfurt Institute for Advanced Studies

Date and time: February 5th, 2025 Wednesday at 8am PST / 11am EST / 4pm GMT / 5pm CET / 12am China

For the zoom links, please join the One World Cryo-EM mailing list.

From Pixels to Bayesian Posteriors: Fast Molecular Conformation Estimation from Cryo-EM with simulation-based inference

Biomolecules are highly dynamic systems. They reorganize between a network of metastable states connected by rare structural intermediates, which is referred to as their conformational ensemble. The ensemble, including the rare intermediate structures, determines biomolecular function in the cell. However, mapping biomolecular conformational ensembles is still an outstanding challenge in both experimental and computational approaches. Cryo-electron microscopy (cryoEM) has emerged as a powerful paradigm for characterizing protein conformational ensembles. However, even though the frozen sample contains information on the entire ensemble, current approaches reconstruct only a few conformations by averaging over many microscopy images. In fact, accurately identifying molecular conformations depicted in a single cryoEM image is still a challenging task. Here, we integrate simulations and probabilistic deep learning to develop the cryoEM simulation-based inference (cryoSBI) framework for inferring molecular conformations and their uncertainties from individual cryoEM images. Given an observed image, cryoSBI enables us to directly estimate the Bayesian posterior using forward model simulations, an embedding network, and a neural posterior estimation framework. CryoSBI is amortized. Training happens only once, after which inference for each experimental image takes only milliseconds to evaluate. Pose and imaging parameters do not have to be estimated, resulting in a high computational speed compared to explicit likelihood methods. For synthetic and experimental data, we could systematically disentangle the molecular conformation from the noisy observation with a confidence interval for the inference and learn about the most relevant features of the observed particles. Our approach paves the way to characterizing entire conformational ensembles from experimental data.

Bio:
Roberto Covino is W3 Professor of Computational Life Science at the Institute of Computer Science at Goethe University Frankfurt, and a Senior Fellow at the Frankfurt Institute for Advanced Studies. His research uses theory, simulation, and statistical modelling of experiments to understand how biomolecular functions emerge from the interplay between structure, dynamics, and complexity. He focuses on understanding the mechanism of key events in proteins and cellular membranes. RC studied physics and theoretical physics at the University of Bologna and obtained his PhD in physics at the University of Trento. He was appointed Professor of AI in Protein Science at Bayreuth University in 2023 and received the call to Goethe University in 2024.

2024/12/11 – Kai (Jack) Zhang

Kai (Jack) Zhang
Assistant Professor, Department of Molecular Biophysics and Biochemistry (MB&B)
Yale University

Date and time: December 11th, 2024 Wednesday at 8am PST / 11am EST / 4pm GMT / 5pm CET / 12am China

For the zoom links, please join the One World Cryo-EM mailing list.

High-resolution in situ structures of mammalian respiratory supercomplexes

Mitochondria are essential for ATP production via oxidative phosphorylation, involving respiratory complexes within the inner membrane. Despite extensive in vitro studies, understanding their mechanisms in physiological environment is challenging due to loss of the native environment during purification. Here, we directly image porcine mitochondria by developing a high-resolution in-situ cryo-electron microscopy technique, which enabled us to determine near-atomic structures of various respiratory supercomplexes in the native membrane.

2024/10/02 – Niels Volkmann

Niels Volkmann
Professor of Electrical and Computer Engineering
UC Santa Barbara

Date and time: October 2th, 2024 Wednesday, at 8am PDT / 11am EDT / 4pm BST / 5pm CEST / 11pm China CST

For the zoom links, please join the One World Cryo-EM mailing list.

High(er) throughput analysis of actin-filament structures in cellular tomograms

The actin cytoskeleton plays a key role in cell migration and morphology in eukaryotic cells. Its diverse architectures enable functions such as protrusion, adhesion, contraction, and retraction. Analyzing cryo-ET data presents challenges due to low signal-to-noise ratios, stemming from weak contrast between biomolecules and the surrounding medium, as well as low electron doses to prevent sample damage. Current methods, such as analyzing vectorized actin filament traces or subtomogram averaging, have provided detailed insights into cytoskeletal structures but do not scale well for large datasets. In this talk, I will present alternative higher-throughput methods we developed for extracting nanoscale actin-filament parameters from hundreds of tomograms.

2024/06/05 – Antonio Martinez Sánchez

Antonio Martinez Sánchez
Professor in Computer Sciences, University of Murcia, Spain

Date and time: June 5th, 2024 Wednesday, at 8am PDT / 11am EDT / 3pm BST / 5pm CEST / 11pm China

For the zoom links, please join the One World Cryo-EM mailing list.

Fast Normalized Cross-Correlation for Template Matching with Rotations

Object detection is a main task in computer vision. Template matching is the reference method for detecting objects with arbitrary templates. However, template matching computational complexity depends on the rotation accuracy, being a limiting factor for large 3D images (tomograms). 

Here, we implement a new algorithm called tensorial template matching, based on a mathematical framework that represents all rotations of a template with a tensor field. Contrary to standard template matching, the computational complexity of the presented algorithm is independent of the rotation accuracy.

Using both, synthetic and real data from cryo-electron tomography, we demonstrate that tensorial template matching is much faster than template matching and has the potential to improve its accuracy.

2024/05/08 – Aaditya Rangan

Aaditya Rangan
Associate Professor, Courant Institute, New York University and Flatiron Institute.

Date and time: May 8th, 2024 Wednesday, at 8am PDT / 11am EDT / 3pm BST / 5pm CEST / 11pm China

Ab-initio Reconstruction from small datasets: using entropy-maximization and principal modes (EMPM).

In this talk I’ll describe a simple algorithm for obtaining a low-resolution model from a small number (~1000) picked-particle images. The algorithm (termed EMPM) is similar to classical alternating-minimization, but with a slight twist: the entropy of the viewing-angle distribution is maximized during the search. This entropy-maximization greatly improves robustness and accuracy, allowing EMPM to outperform many existing ab-initio reconstruction algorithms. Because of its low computational overhead, this algorithm could facilitate statistical strategies such as bootstrapping and cross-validation, along with the rapid assessment of particle quality.

2024/04/10 Yuntao Liu

Yuntao Liu

Postdoctoral scholar, California NanoSystems Institute, UCLA

Date and time: Apr 10th, 2024
Wednesday, at 8am PST / 11am EDT / 3pm GMT / 5pm CET / 11pm China

For the zoom and gather.town links, please join the One World Cryo-EM mailing list.

Single-particle IsoNet: an AI-based method to overcome the “preferred” orientation problem

While advances in single-particle cryoEM have enabled the structural determination of macromolecular complexes at atomic resolution, particle orientation bias (the so-called “preferred” orientation problem) remains a complication for most specimens. Existing solutions have relied on biochemical and physical strategies applied to the specimen and are often complex and challenging. Here, we develop spIsoNet, an end-to-end self-supervised deep-learning software to address the preferred orientation problem. Using preferred-orientation views to recover molecular information in under-sampled views, spIsoNet improves both angular isotropy and particle alignment accuracy during 3D reconstruction. We demonstrate spIsoNet’s capability of generating near-isotropic reconstructions from representative biological systems with limited views, including ribosomes, β-galactosidases, and a previously intractable hemagglutinin trimer dataset. spIsoNet can also be generalized to improve map isotropy and particle alignment of preferentially oriented molecules in subtomogram averaging. Therefore, without additional specimen-preparation procedures, spIsoNet provides a general computational solution to the preferred orientation problem. 

2024/03/06 Quanquan Gu

Quanquan Gu

Head of AI for Drug Design, ByteDance Research

Date and time: Mar 6th, 2024
Wednesday, at 8am PST / 11am EST / 4pm GMT / 5pm CET / 12am China

For the zoom and gather.town links, please join the One World Cryo-EM mailing list.

CryoSTAR: Leveraging Structural Prior and Constraints for Cryo-EM Heterogeneous Reconstruction

Resolving conformational heterogeneity in cryo-electron microscopy (cryo-EM) datasets remains a significant challenge in structural biology. Previous methods have often been restricted to working exclusively on volumetric densities, neglecting the potential of incorporating any pre-existing structural knowledge as prior or constraints. In this talk, I will present a novel methodology, cryoSTAR, that harnesses atomic model information as structural regularization to elucidate such heterogeneity. Our method uniquely outputs both coarse-grained models and density maps, showcasing the molecular conformational changes at different levels. Validated against four diverse experimental datasets, spanning large complexes, a membrane protein, and a small single-chain protein, our results consistently demonstrate an efficient and effective solution to conformational heterogeneity with minimal human bias. By integrating atomic model insights with cryo-EM data, cryoSTAR represents a meaningful step forward, paving the way for a deeper understanding of dynamic biological processes. This is joint work with Yilai Li, Yi Zhou, Jing Yuan and Fei Ye.

More graphics are available at: https://bytedance.github.io/cryostar/.

2023/12/06 Marc A. Gilles

Marc A. Gilles

Postdoctoral Research Associate, Program in Applied & Computational Mathematics, Princeton University

Date and time: Dec 6th, 2023
Wednesday, at 8am PST / 11am EST / 4pm GMT / 5pm CET / 12am China

For the zoom and gather.town links, please join the One World Cryo-EM mailing list.

A Bayesian Framework for Cryo-EM Heterogeneity Analysis using Regularized Covariance Estimation

Proteins and the complexes they form are central to nearly all cellular processes. Their flexibility, expressed through a continuum of states, provides a window into their biological functions. Cryogenic-electron microscopy (cryo-EM) is an ideal tool to study these dynamic states as it captures specimens in non-crystalline conditions and enables high-resolution reconstructions. However, analyzing the heterogeneous distribution of conformations from cryo-EM data is challenging. Current methods face issues such as a lack of explainability, overfitting caused by lack of regularization, and a large number of parameters to tune; problems exacerbated by the lack of proper metrics to evaluate or compare heterogeneous reconstructions. To address these challenges, we present RECOVAR, a white-box method based on principal component analysis (PCA) computed via regularized covariance estimation that can resolve intricate heterogeneity with similar expressive power to neural networks with significantly lower computational demands.

2023/11/01 Bronwyn Lucas

Bronwyn Lucas

Assistant Professor of Biochemistry, Biophysics and Structural Biology, UC Berkeley

Date and time: Nov 1st, 2023
Wednesday, at 8am PDT / 11am EDT / 3pm GMT / 5pm CET / 11pm China

For the zoom and gather.town links, please join the One World Cryo-EM mailing list.

Baited reconstruction for high-resolution in situ structure determination without template bias

Cryogenic electron microscopy (cryo-EM) has revolutionized structural biology, rapidly increasing the number of available molecular structures. Because of this, as well as advances in structure prediction, the focus of structural biology has begun to shift to studying macromolecular structures in their native cellular environment. A dominant feature of cryo-EM images is shot noise, making the identification of small particles of interest difficult. This is further compounded by structural noise if these particles are imaged against a background of other molecules, such as inside a cell. High-resolution template matching is currently the most effective method for visual proteomics in 2D and 3D. We have shown that 2D template matching (2DTM) can be used to localize complexes with high precision, even in the presence of cellular background. Once localized, these particles may be averaged together in 3D reconstructions; however, in both 2D and 3D, regions included in the template may suffer from template bias, leading to inflated resolution estimates and making the interpretation of high-resolution features unreliable.

I will describe our new work where we evaluated conditions that minimize template bias and show that molecular features not present in the template can be reconstructed at high resolution from targets found by 2DTM, extending prior work at low-resolution. Moreover, I will describe a quantitative metric for template bias to aid the interpretation of 3D reconstructions calculated with particles localized using high-resolution templates and fine angular sampling. Baited reconstruction allows us to retain the benefits of precise particle localization in situ with high-resolution template matching while accounting for and minimizing overfitting.

2023/05/03 Amit Moscovich

Amit Moscovich

Assistant Professor, Department of Statistics and Operations Research, Tel Aviv University

Date and time: May 3rd, 2023
Wednesday, at 8am PDT / 11am EDT / 4pm BST / 5pm CEST / 11pm China

For the zoom and gather.town links, please join the One World Cryo-EM mailing list.

Tools for heterogeneity in cryo-EM: manifold learning, disentanglement and optimal transport

Despite much recent progress, the reconstruction and analysis of macromolecules with continuous conformational heterogeneity remains a key challenge.  In this talk we present two promising tools that can aid in various stages of the reconstruction and analysis pipelines.

In the first part of the talk, the notion of manifold factorization and disentanglement is presented.  This is an approach to dimensionality reduction where different aspects of the data are assigned separate coordinates.  It has potential applications both for the reconstruction of single-particle cryo-EM samples with continuous heterogeneity and for the analysis of the reconstructed volumetric datasets.

In the second part of the talk, we will discuss the potential applications of optimal transport for cryo-EM, in particular for the analysis of heterogeneous samples, as well as for class-averaging and particle picking.  Optimal transport metrics are closely related to physical motion, making them a natural choice for many of the core problems in cryo-EM.  Historically, computational bottlenecks have limited the applicability of optimal transport.  However, recent advances in computational optimal transport have yielded fast approximation schemes that can be readily used for the analysis of high-resolution images and volumetric arrays.