CryoAtom Cryo-EM Model Builder
CryoAtom builds atomic models from cryo-EM maps using local attention and 3D rotary position embedding, improving model completeness and speed while lowering resolution requirements.
CryoAtom is an approach for de novo model building from cryogenic electron microscopy (cryo-EM) density maps. It leverages advancements from AlphaFold2, replacing global attention with a local attention mechanism enhanced by a novel 3D rotary position embedding. This allows CryoAtom to produce more complete atomic models, lower the required map resolution, and significantly accelerate the modeling process compared to existing methods.
Estimated resource cost: 0.000527
Categories: Structure Prediction & Folding
Key Capabilities
- Leverages a local attention mechanism and 3D rotary position embedding for improved accuracy.
- Produces more complete models and reduces the resolution requirement for cryo-EM maps.
- Accelerates the modeling process, enabling the construction of large, 100+ protein complexes in hours.
- Accurately distinguishes between paralog sequences, even in noisy map regions.
- Detects and models previously uncharacterized proteins and structural extensions.
- Captures minor conformational changes within mega-Dalton complexes.
Runtime Statistics
| Metric | Value |
|---|---|
| runtime_mean | 985 |
| runtime_median | 621 |
| runtime_std | 894 |
| runtime_90th_percentile | 1934 |
| runtime_max | 4486 |
Similar Tools
- CryoSAMU
- DeepEMhancer
- AFcluster
- AlphaFold2
Ready to submit your job?
Review your configuration, then confirm the estimated credit cost before you run the job. Note that credit estimates are not guaranteed and runtime can vary depending on inputs and settings.
Estimated Credits: 0.000527
Invite-only, limited-time access. Please contact ztang@getantibody.com.
Invite-only, limited-time access. Please contact ztang@getantibody.com.