Sparse Fourier Backpropagation in Cryo-EM Reconstruction

Authors: Dari Kimanius, Kiarash Jamali, Sjors Scheres

NeurIPS 2022 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We achieve improved results on a simulated data set and at least equivalent results on an experimental data set when compared to the coordinate-based approach, while also substantially lowering computational cost.
Researcher Affiliation Academia Dari Kimanius Kiarash Jamali Sjors H.W. Scheres MRC Laboratory of Molecular Biology {dari, kjamali, scheres}@mrc-lmb.cam.ac.uk
Pseudocode No The paper does not contain any pseudocode or clearly labeled algorithm blocks.
Open Source Code Yes The code used in this paper is available at github.com/dkimanius/sbackprop, together with the instructions for compiling and running it.
Open Datasets Yes This data set, along with the PDB files used to create it and the refinement results, can be found at zenodo.org/record/7182156 (DOI:10.5281/zenodo.7182156).
Dataset Splits No The paper does not explicitly provide training, validation, or test dataset splits. It mentions the total number of images in the datasets but not how they were partitioned for training or evaluation.
Hardware Specification Yes All calculations were performed on an NVIDIA RTX 3090.
Software Dependencies No The paper mentions "Py Torch [21] with custom CUDA kernels" but does not specify version numbers for PyTorch or CUDA, which are necessary for reproducible software dependencies.
Experiment Setup Yes We trained our VAE with 8 latent dimensions for 50 epochs... We used 10 latent dimensions for this data set.