Super-Resolution Cryo-EM Maps With 3D Deep Generative Networks
Subramaniya, S. R. M. V.; Terashi, G.; Kihara, D.
Show abstract
An increasing number of biological macromolecules have been solved with cryo-electron microscopy (cryo-EM). Over the past few years, the resolutions of density maps determined by cryo-EM have largely improved in general. However, there are still many cases where the resolution is not high enough to model molecular structures with standard computational tools. If the resolution obtained is near the empirical border line (3-4 [A]), a small improvement of resolution will significantly facilitate structure modeling. Here, we report SuperEM, a novel deep learning-based method that uses a three-dimensional generative adversarial network for generating an improved-resolution EM map from an experimental EM map. SuperEM is designed to work with EM maps in the resolution range of 3 [A] to 6 [A] and has shown an average resolution improvement of 1.0 [A] on a test dataset of 36 experimental maps. The generated super-resolution maps are shown to result in better structure modelling of proteins.
Matching journals
The top 4 journals account for 50% of the predicted probability mass.
Similar papers in this journal
Similar papers in this journal
- DomainFit: Identification of Protein Domains in cryo-EM maps at Intermediate Resolution using AlphaFold2-predicted Models 97%
- GemSpot: A Pipeline for Robust Modeling of Ligands into CryoEM Maps 95%
- Cost-benefit analysis of cryogenic electron tomography subtomogram averaging of chaperonin MmCpn at near atomic resolution. 95%
Similar papers in this journal
- Predicting structures of large protein assemblies using combinatorial assembly algorithm and AlphaFold2 96%
- AlphaFold predictions are valuable hypotheses, and accelerate but do not replace experimental structure determination 96%
- Outcomes of the 2019 EMDataResource model challenge: validation of cryo-EM models at near-atomic resolution 96%
Similar papers in this journal
"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.