cryoSPHERE: Single-Particle HEterogeneous REconstruction from cryo EM

Authors: Gabriel Claude Jean Ducrocq, Lukas Grunewald, Sebastian Westenhoff, Fredrik Lindsten

ICLR 2025 | Venue PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We demonstrate this with two synthetic datasets featuring varying levels of noise, as well as two real dataset. We show that cryo SPHERE is very resilient to the high levels of noise typically encountered in experiments, where we see consistent improvements over the current state-of-the-art for heterogeneous reconstruction.
Researcher Affiliation Academia Gabriel Ducrocq Division of Statistics and Machine Learning Link oping University, Link oping, Sweden EMAIL Lukas Grunewald Department of Chemistry Uppsala University, Uppsala, Sweden EMAIL Sebastian Westenhoff Department of Chemistry Uppsala University, Uppsala, Sweden EMAIL Fredrik Lindsten Division of Statistics and Machine Learning Link oping University, Link oping, Sweden EMAIL
Pseudocode No The paper does not contain a clearly labeled pseudocode or algorithm block.
Open Source Code Yes The implementation of the model is available on github 1. 1https://github.com/Gabriel-Ducrocq/cryo SPHERE
Open Datasets Yes We now demonstrate that cryo SPHERE is applicable to real data as well as large proteins. We run cryo SPHERE on EMPIAR-10180 Plaschka et al. (2017)... We now tackle the recently published EMPIAR-12093 (B odizs et al., 2024).
Dataset Splits No The paper mentions datasets but does not explicitly provide training/test/validation splits. For the real datasets, it uses the entire EMPIAR-10180 and EMPIAR-12093 datasets without specifying a split.
Hardware Specification Yes We train cryo SPHERE with Nsegm = 25, cryo Star, and cryo DRGN for 24 hours each, using the same single GPU. ... Our computations were enabled by the Berzelius resource at the National Supercomputer Centre, provided by the Knut and Alice Wallenberg Foundation.
Software Dependencies No The paper mentions software like AlphaFold, RELION, cryoSPARC, cryoDRGN, cryoStar, GROMACS, and PLUMED, but does not specify version numbers for these or other key dependencies used in their own implementation.
Experiment Setup Yes We train cryo SPHERE with Nsegm = 25, cryo Star, and cryo DRGN for 24 hours each, using the same single GPU. ... We set τ = 20 in the experiment section.