Biophysical Society Thematic Meeting | Trieste 2024

Emerging Theoretical Approaches to Complement Single-Particle Cryo-EM

Thursday Speaker Abstracts

TOWARDS QUANTITATIVE RECOVERY OF PROBABILITY DENSITIES FROM CRYO-EM Jake Moomaw 1 ; Joshua Rhodes 1 ; Diego Sanchez 1 ; Jeffrey Xiang 1 ; Erik H Thiede 1 ; 1 Cornell University, Chemistry and Chemical BIology, Ithaca, NY, USA Many biological molecules are structurally heterogeneous. In principle, cryo-electron microscopy (cryoEM) gives us the tools to extract this heterogeneity since the snap-freezing process traps biomolecules in conformations close to the ones they adopt in solution. This means that, modulo the effects of freezing on the conformational ensemble, we should be able to quantitatively recover conformational probabilities from cryo-EM. Unfortunately, cryo-EM’s low signal-to-noise ratio, as well as the complexity of proteins’ conformational landscapes, makes this a challenging task. Prior work that attempts to quantitatively recover the probabilities of protein conformations often finds itself hamstrung by the computational expense of comparing conformational hypotheses with images and of generating conformational hypotheses through simulation. Here, we discuss our progress towards addressing these problems. We first discuss our work in using fast algorithms to accelerate comparisons of cryo-EM images with conformations. By filtering out comparisons that don’t effect the final result, we can improve the scaling from linear to logarithmic in the size of the hypothesis-space, leading to substantial improvements in speed. Next, we discuss our initial work in using of generative machine learning to capture the conformational ensemble. Using generative machine learning, we can generate physically-meaningful conformations efficiently and without requiring long molecular simulations. UNCOVERING PROTEIN ENSEMBLES: AUTOMATED MULTICONFORMER MODEL BUILDING FOR CRYO-EM PROTEINS, NUCLEIC ACIDS, SOLVENTS, AND LIGANDS James Fraser 1 ; 1 University of California, San Francisco, San Francisco, California, USA

No Abstract

27