Home Research Services
Online Services
●D-I-TASSER ●DMFold ●LOMETS3 ●DeepMSA2 ●HPmod ●BioLiP2 ●US-align ●InterLabelGO+ ●StarFunc ●ShapeME ●StoPred

campaign Release and Update Notice

Updated May 30, 2026
  • A new StoPred model collection has been deployed.
  • Model selection is available so users can choose the current default complex model, the manuscript complex model, or the single-sequence unknown model. Dates in model names indicate the training cutoff date.
  • The single-sequence unknown model is intended for one protein chain when the user does not know whether it is a monomer or a homomer.
  • A separate download page is available for container images, model releases, and model naming documentation.

info About StoPred

StoPred is a deep learning-based tool that predicts the stoichiometry of protein complexes from sequence alone, using a combination of protein language model embeddings and a graph attention neural network.

StoPred can handle both homomeric and heteromeric complexes, without requiring template assemblies or predefined stoichiometry. The model outputs the most probable stoichiometry and can suggest a small set of candidates to guide downstream structure prediction.

StoPred achieves state-of-the-art performance compared to Template-based and AlphaFold ranking score-based methods on large-scale benchmark datasets.
Example output

rule Model Use Cases

  • Default complex model: use this when the input proteins are expected to form a complex. This is the standard StoPred setting and assumes the submitted chains are possible binding partners. The date indicates the training cutoff date.
  • Manuscript complex model: use this to reproduce predictions from the StoPred manuscript model. The date indicates the training cutoff date.
  • Single-sequence unknown model: use this only when one protein chain is submitted and you do not know whether it is a monomer or a homomer. This model is intended to estimate likely homo-oligomeric stoichiometry from one protein sequence alone. The date indicates the training cutoff date.
  • Multi-chain FASTA: submit one FASTA record per chain when the candidate complex members are known. Each chain should have a unique FASTA header.

upload_file Submit a Protein Complex

Choose a complex model when submitted chains are expected to form a complex. Choose the single-sequence unknown model only when one protein chain is submitted and you do not know whether it is a monomer or a homomer.
Paste your multi-chain FASTA or upload a file. Each chain should have a unique header.
The default model assumes the submitted sequence(s) represent a protein complex.

download Download

download
Download Page
Container image, software package, model releases, and model naming documentation
Open Downloads
storage
Model Releases
Available StoPred model archives and model naming documentation
View Models

menu_book Citation

Liu, Q., Peng, C., Zheng, W., Zhang, C., & Freddolino, L. (2025). StoPred: Accurate Stoichiometry Prediction for Protein Complexes Using Protein Language Models and Graph Attention. bioRxiv. https://doi.org/10.1101/2025.10.20.683515

FreddolinoLabumich.edu | 1150 W. Medical Center Dr, Ann Arbor, MI 48109-0600