Skip to content

Working with EMDB

This section describes recommended practices for accessing and working with data from the Electron Microscopy Data Bank (EMDB), with a focus on scalable, reliable, and reproducible workflows.

For large-scale data retrieval from EMDB, FTP-based access is strongly recommended over HTTPS.
FTP provides significantly better stability and throughput, especially when downloading large numbers of entries or full-resolution density maps. EMDB maintains multiple geographically distributed mirrors. Users should select the mirror closest to their location to maximize download performance.

For users located in East Asia (e.g. China, Japan, Korea), the PDBj mirror provides substantially improved stability and download speed compared to the primary EBI servers.

Example: Downloading Entry EMD-0153

# Using wget
wget https://ftp.pdbj.org/pub/emdb/structures/EMD-0153/map/emd_0153.map.gz

Programmatic Dataset Selection via EMDB Search API

Before downloading large datasets, we strongly recommend using the EMDB Search API to programmatically identify entries of interest.
This enables efficient filtering based on metadata and avoids unnecessary data transfer.

Common filtering criteria include:

  • Resolution range
  • Molecular weight
  • Structure determination method (e.g. single-particle cryo-EM)
  • Entry status

Example query (CSV output)

https://www.ebi.ac.uk/emdb/api/search/database:EMDB AND current_status:[* TO *] AND assembly_molecular_weight:{0 TO 10000000] AND resolution:[0 TO 3} AND structure_determination_method:"singleparticle"?rows=1000000&wt=csv&download=false&fl=emdb_id,structure_determination_method,resolution,resolution_method,assembly_molecular_weight,cpx_name

This interface allows users to: - programmatically retrieve metadata, - construct reproducible dataset selection pipelines, - selectively download only relevant EMDB entries.

Acknowledgments

We thank Kyle Morris for his contributions and insights regarding EMDB data access and workflow recommendations.