This page keeps track of software that is actively developed or maintained by people at Rostlab. Please take a look at the repositories or linked documentation, to get further information or help.

Software

• ProtSpace: Explore, analyze, understand your protein space.
  GitHub: https://github.com/tsenoner/protspace_web 
  Web: https://protspace.app/
  Paper: https://doi.org/10.1016/j.jmb.2025.168940  
• biocentral: Embedding-based protein predictions.
  GitHub: https://github.com/biocentral/biocentral 
  Web: https://biocentral.rostlab.org/ 
  Paper: https://doi.org/10.1016/j.jmb.2026.169673
• TMvisDB: Provides per-residue transmembrane topology annotations for all proteins in AlphaFold DB (~ 200 million proteins, September '22) predicted as transmembrane proteins (~ 46 million).
  GitHub: https://github.com/t03i/TMVisDB 
  Web: https://tmvisdb.rostlab.org/ 
  Paper: https://doi.org/10.1016/j.jmb.2025.168997
• biotrainer: Biological prediction models made simple.
  GitHub: https://github.com/sacdallago/biotrainer 

Models

Protein Language Models:

• ProtT5: State-of-the-art protein language model.
  GitHub: https://github.com/agemagician/ProtTrans 
  Paper: https://doi.org/10.1109/tpami.2021.3095381
• ProstT5: Bilingual Language Model for Protein Sequence and Structure.
  GitHub: https://github.com/mheinzinger/ProstT5 
  Paper: https://doi.org/10.1093/nargab/lqae150 

Protein Prediction Models:

• VespaG: Expert-Guided Protein Language Models enable Accurate and Blazingly Fast Fitness Prediction.
  GitHub: https://github.com/JSchlensok/VespaG/ 
  Paper: https://doi.org/10.1093/bioinformatics/btae621
• TMbed: Transmembrane proteins predicted through Language Model embeddings.
  GitHub: https://github.com/BernhoferM/TMbed 
  Paper: https://doi.org/10.1186/s12859-022-04873-x
• SETH: ProtT5 (Transformer) embeddings used for residue wise disorder prediction in proteins.
  GitHub: https://github.com/DagmarIlz/SETH 
  Paper: https://doi.org/10.3389/fbinf.2022.1019597
• LightAttention: Using Transformer protein embeddings with a linear attention mechanism to make SOTA de-novo predictions for the subcellular location of proteins.
  GitHub: https://github.com/HannesStark/protein-localization 
  Paper: https://doi.org/10.1093/bioadv/vbab035 

All models listed above are available via our web service.

Datasets

• PBC: Protein (language model) Benchmarking Collection. Automated benchmarking available via biotrainer-autoeval.
  GitHub: https://github.com/Rostlab/pbc