Research map / 3D vision / Face and shape
CryoDRGN2: Ab Initio Neural Reconstruction of 3D Protein Structures From Real Cryo-EM Images
Ellen D. Zhong, Adam Lerer, Joseph H. Davis, Bonnie Berger
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| use cryoDRGN2 pose search instead of branch and bound at 30 degree grid | cryoDRGN · 80S, 1000 images | alignment time: 0:01:32 -> 0:00:23 | Improved |
| use cryoDRGN2 pose search instead of branch and bound at 30 degree grid | cryoDRGN · 80S, 1000 images | mean rotation error: 0.691 -> 0.643 | Improved |
| refine the pose grid from 30 deg/2.8 pix to 15 deg/1.4 pix (ablation) | cryoDRGN2 · 80S, 1000 images | mean rotation error: 0.643 -> 0.004 | Improved |
| refine the pose grid from 30 deg/2.8 pix to 15 deg/1.4 pix (ablation) | cryoDRGN2 · 80S, 1000 images | alignment time: 0:00:23 -> 0:00:52 | Hurt |
| use exhaustive pose search instead of learned (pose VAE) or gradient-based pose optimization | cryoDRGN2 · Synthetic Hand noiseless | median rotation error: 6.66 -> 0.0003 | Improved |
| add model reset and continued pose-search refinement (cryoDRGN2+r) (ablation) | cryoDRGN2 · RAG12 | median rotation error: 0.0386 -> 0.0226 | Improved |
| add model reset and continued pose-search refinement (cryoDRGN2+r) (ablation) | cryoDRGN2 · RAG12 | mean rotation error: 3.4254 -> 3.3730 | Improved |
| add model reset and continued pose-search refinement (cryoDRGN2+r) (ablation) | cryoDRGN2 · 80S | median rotation error: 0.0008 -> 0.0008 | No clear effect |
| add model reset and continued pose-search refinement (cryoDRGN2+r) (ablation) | cryoDRGN2 · Spliceosome | median rotation error: 0.0046 -> 0.0044 | No clear effect |
| use cryoDRGN2 instead of cryoDRGN-BNB | cryoDRGN · RAG12 | median rotation error: 4.6371 -> 0.0386 | Improved |
| use cryoDRGN2 instead of cryoDRGN-BNB | cryoDRGN · Spliceosome | median rotation error: 0.1854 -> 0.0046 | Improved |
| use cryoDRGN2 instead of cryoSPARC on noisy synthetic data | cryoDRGN2 · Synthetic Hand SNR=0.1 | median rotation error: 0.071 -> 0.027 | Improved |
Later papers that compared with this method
4 later accepted papers report CryoDRGN2 as a baseline in their results tables; 2 of them report a better number on most of the results they share with it.
- CryoFastAR: Fast Cryo-EM Ab initio Reconstruction Made Easy (ICCV 2025 · better on 5 of 6 shared results)
- Mixture of neural fields for heterogeneous reconstruction in cryo-EM (NeurIPS 2024 · better on 4 of 4 shared results)
- Amortized Inference for Heterogeneous Reconstruction in Cryo-EM (NeurIPS 2022 · better on 5 of 11 shared results)
- CryoAI: Amortized Inference of Poses for Ab Initio Reconstruction of 3D Molecular Volumes from Real Cryo-EM Images (ECCV 2022 · better on 0 of 3 shared results)
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- cryoSPHERE: Single-Particle HEterogeneous REconstruction from cryo EM (ICLR 2025 · similarity 0.92)
- CryoSPIN: Improving Ab-Initio Cryo-EM Reconstruction with Semi-Amortized Pose Inference (NeurIPS 2024 · similarity 0.91)
- Mixture of neural fields for heterogeneous reconstruction in cryo-EM (NeurIPS 2024 · similarity 0.91)
- CryoBench: Diverse and challenging datasets for the heterogeneity problem in cryo-EM (NeurIPS 2024 · similarity 0.90)
- Reconstructing Heterogeneous Biomolecules via Hierarchical Gaussian Mixtures and Part Discovery (NeurIPS 2025 · similarity 0.89)
- Multiscale guidance of protein structure prediction with heterogeneous cryo-EM data (NeurIPS 2025 · similarity 0.89)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
