Proteomes2Structs is a data acquisition tool for reproducible, proteome‑scale downloads of AlphaFold DB structures and metadata. It is designed for downstream pipelines that require consistent directory layouts, structured metadata, and reliable parallel downloads across thousands of proteins. At present, only UniProt proteomes are supported.
Download structures for E. coli:
proteomes2structs --cif "UP000000625" dataClone the repository:
git clone https://github.com/CielResearch/Proteomes2Structs.git
cd Proteomes2StructsMake the wrapper executable:
chmod +x bin/proteomes2structs(Optional) Add the tool to your PATH:
export PATH="$PWD/bin:$PATH"You can now run:
proteomes2structs --helpOnce the Bioconda recipe is accepted, you will be able to install Proteomes2Structs with:
conda install -c bioconda proteomes2structsUsage: proteomes2structs [options] "PROTEOME_LIST" OUTPUT_DIR
Example: proteomes2structs --cif "UP000000625 UP000007256" ../data
| Positional Argument | Example | Notes |
|---|---|---|
| "PROTEOME_LIST" | "UP000000625 UP000007256" | Quoted space-separated UniProt proteome accession/s |
| OUTPUT_DIR | ../data | Directory where downloaded files will be stored |
| Category | Flag/Option | Default | Notes |
|---|---|---|---|
| File format | --cif |
Download .mmCIF/.cif files | |
| File format | --pdb |
Download .pdb files | |
| Parallelism | --threads |
12 | Number of download threads |
| Mode | --mode=download |
Run proteomes2structs in download mode (default mode) | |
| Other | --keep-fasta |
Do not automatically delete FASTA files | |
| Other | --sui |
5 | Status update interval in minutes |
- At least one file format flag must be enabled (
--cifor--pdb).
OUTDIR/
proteome1/
json/ # AFDB metadata snapshots per protein
structures/ # .cif and/or .pdb files per protein
logs/
failures_fasta.txt
failures_metadata.txt
failures_structures.txt
metadata.json # summary of timestamps, success/failure counts, taxa ID, mode
proteome2/
...
proteomeK/
Proteomes2Structs downloads structures from AlphaFold DB for any UniProt proteome with AFDB coverage. Note that AFDB does not host most viral proteomes (except SARS‑CoV‑2), so viral proteomes will produce empty metadata and no structure files.
Proteomes2Structs is designed for reproducible, large‑scale structural downloads. When tested on ~20,000 proteins across six proteomes (CHANGELOG.md v0.3.0a1), the pipeline achieved:
- ~17 minutes total runtime on a home laptop (WSL2, 16 threads)
- 0% structure download failures (excluding viral proteomes, which AFDB does not host)
- ~0.4% metadata download failure rate, all due to missing AFDB metadata rather than pipeline errors
- Clean handling of unsupported proteomes, such as bacteriophage λ
These numbers reflect AFDB’s upstream coverage rather than issues with Proteomes2Structs. Proteomes2Structs has not been tested on HPC environments, but runtimes are likely faster.
Please open an issue on GitHub for bug reports or feature requests.
Bertoni, D., Tsenkov, M., Magana, P., Nair, S., Pidruchna, I., Querino Lima Afonso, M., Midlik, A., Paramval, U., Lawal, D., Tanweer, A., Last, M., Patel, R., Laydon, A., Lasecki, D., Dietrich, N., Tomlinson, H., Žídek, A., Green, T., Kovalevskiy, O., … Velankar, S. (2026). AlphaFold Protein Structure Database 2025: A redesigned interface and updated structural coverage. Nucleic Acids Research, 54(D1), D358–D362. https://doi.org/10.1093/nar/gkaf1226
The UniProt Consortium. (2025). UniProt: The Universal Protein Knowledgebase in 2025. Nucleic Acids Research, 53(D1), D609–D617. https://doi.org/10.1093/nar/gkae1010
Baumann, C. I.-L. (2026). Proteomes2Structs [Shell]. Zenodo. https://doi.org/10.5281/zenodo.21850698
@software{baumann_proteomes2structs_2026, author = {Baumann, Ciel Ivy-Lee}, title = {Proteomes2Structs}, doi = {10.5281/zenodo.21850698}, url = {https://github.com/CielResearch/Proteomes2Structs}, year = {2026} }
This tool is licensed under Creative Commons Attribution–NonCommercial 4.0 International (CC BY‑NC 4.0). Commercial use and patenting are prohibited. Academic use is permitted with citation.