PIPENN-EMB: ensemble net and protein embeddings generalise protein interface prediction beyond homology
Open accessPreprint — published as 10.1038/s41598-025-88445-y · Paper page
The DOI matches its Crossref record; 2 of the 2 references it cites were found in a registry or an index.
Abstract
Protein interactions are crucial for understanding biological functions and disease mechanisms, but predicting these remains a complex task in computational biology. Increasingly, Deep Learning models are having success in interface prediction. This study presents PIPENN-EMB which explores the added value of using embeddings from the ProtT5-XL protein language model. Our results show substantial improvement over the previously published PIPENN model for protein interaction interface prediction, reaching an MCC of 0.313 vs. 0.249, and AUC-ROC 0.800 vs. 0.755 on the BIO_DL_TE test set. We furthermore show that these embeddings cover a broad range of ‘hand-crafted’ protein features in ablation studies. PIPENN-EMB reaches state-of-the-art performance on the ZK448 dataset for protein-protein interface prediction. We showcase predictions on 25 resistance-related proteins from Mycobacterium tuberculosis . Furthermore, whereas other state-of-the-art sequence-based methods perform worse for proteins that have little recognisable homology in their training data, PIPENN-EMB generalises to remote homologs, yielding stable AUC-ROC across all three test sets with less than 30% sequence identity to the training dataset, and even to proteins with less than 15% sequence identity. Availability: Webserver, source code and datasets at www.ibi.vu.nl/programs/pipennemb/
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Figures




Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
Highly accurate protein structure prediction with AlphaFold
Jumper, J., Evans, R., Pritzel, A., Green, T., Figurnov, M., Ronneberger, O., Tunyasuvunakool, K., Bates, R., Žídek, A., Potapenko, A., et al. (2021) · Nature · doi:10.1038/s41586-021-03819-2 · Cited by 43,764
Modeling aspects of the language of life through transfer-learning protein sequences
Heinzinger, M., Elnaggar, A., Wang, Y., Dallago, C., Nechaev, D., Matthes, F., & Rost, B. (2019) · BMC Bioinformatics · doi:10.1186/s12859-019-3220-8 · Cited by 606 · cited 1×
“Protein Language Models (PLMs) extend the concepts of Natural Language Processing (NLP) language models to the realm of protein sequences: amino acids are treated as tokens (cf. words in NLP), and entire protein sequences are regarded as sentences 20,35 .”
Background
Research works printed but not linked: 51
- 10.1073/pnas.93.1.13 · DOI · Paper page
- 10.1074/mcp.R113.027771 · DOI · Paper page
- 10.1038/s41588-020-00774-y · DOI · Paper page
- 10.2174/138161212802651661 · DOI
- 10.1038/nrd.2016.29 · DOI · Paper page
- 10.1016/J.CEB.2023.102291 · DOI
- 10.1038/s41598-021-81132-8 · DOI · Paper page
- 10.1093/bib/bbp021 · DOI · Paper page
- 10.1016/j.febslet.2015.10.003 · DOI · Paper page
- 10.1093/bib/bbx022 · DOI · Paper page
- 10.1093/bioinformatics/005 · doi:10.1093/bioinformatics/005
- 10.1093/bioinformatics/btac071 · DOI
- 10.1002/prot.26609 · DOI · Paper page
- 10.1038/s41586-021-03828-1 · DOI · Paper page
- 10.1002/ADVS.202102592 · DOI · Paper page
- 10.1093/bioinformatics/btab761 · DOI · Paper page
- 10.1038/s41591-021-01533-0 · DOI · Paper page
- 10.1038/s41592-021-01365-3 · DOI · Paper page
- 10.1101/864405 · DOI
- 10.1038/s41598-020-80786-0 · DOI · Paper page
- 10.1038/s41598-021-00311-9 · DOI
- 10.1038/s41598-022-13951-2 · DOI
- 10.1038/s41598-022-19608-4 · DOI · Paper page
- 10.1093/bioinformatics/btac020 · DOI · Paper page
- 10.1093/bioinformatics/btaa701 · DOI · Paper page
- 10.1109/TPAMI.2021.3095381 · DOI · Paper page
- 10.1073/pnas.2016239118 · DOI · Paper page
- 10.34133/research.0240 · DOI
- Cui, Y. , Dong, Q. & Hong, D. Predicting protein-ligand binding residues with deep convolutional neural networks. BMC Bioinforma. 20 (2019). · DOI · Paper page
- 10.1016/j.engappai.2022.105151 · DOI · Paper page
- 10.1016/J.CSBJ.2020.10.017 · DOI
- 10.3389/FMOLB.2021.619403/BIBTEX · doi:10.3389/fmolb.2021.619403/bibtex
- 10.1093/bioinformatics/btm098 · DOI · Paper page
- 10.1038/s41467-018-02974-x · DOI · Paper page
- 10.1186/s13073-018-0611-9 · DOI · Paper page
- 10.1093/nar/gkab988 · DOI · Paper page
- 10.1093/BIOINFORMATICS/BTAD738 · DOI · Paper page
- 10.1002/prot.26615 · DOI
- 10.1038/s41467-022-35677-5 · DOI · Paper page
- 10.1093/bioinformatics/btae012 · DOI
- 10.1371/journal.pcbi.1010669 · DOI
- 10.1006/jmbi.1990.9999 · doi:10.1006/jmbi.1990.9999 · Paper page
- 10.1186/1471-2105-10-421 · DOI · Paper page
- 10.1093/bioinformatics/btaa750 · DOI · Paper page
- 10.1093/bioinformatics/btz324 · DOI · Paper page
- 2014, SPRINGS: Prediction of Protein-Protein Interaction Sites Using Artificial Neural Networks, J Proteom & Comput. Biol, 1, 7
- 10.1016/j.csbj.2021.04.034 · DOI
- 10.1093/NARGAB/LQAC043 · doi:10.1093/nargab/lqac043
- 10.1101/2022.12.15.519894 · DOI
- 10.1038/s41586-021-03819-2 · DOI · Paper page
- 10.1038/s41598-020-75467-x · DOI · Paper page
Metadata
- Authors
- Thomas, D. P. G., Garcia Fernandez, C. M., Haydarlou, R., & Feenstra, K. A.
- Affiliations
- Vrije Universiteit Amsterdam (NL)
- Corresponding author
- Haydarlou, Reza
- Feenstra, K. Anton
- Venue
- Scientific Reports
- Publisher
- openRxiv
- Published
- 2024-11-02
- Volume
- 15
- Issue
- 1
- Pages
- 4391-4391
- ISSN
- 2045-2322 (ISSN-L)
- PMC
- PMC11799512
- OpenAlex
- W4407153268
- Cited by
- 4
- References
- 2
- Citations by year
- 2026: 4
- FWCI
- 1.20
- Type
- journal article
- Language
- en
- Open access
- gold
- doi.org
- Licence
- Locations
- Scientific Reports (publishedVersion, cc-by-nc-nd)
- bioRxiv (Cold Spring Harbor Laboratory) (acceptedVersion, cc-by-nc)
- Research Square (acceptedVersion, cc-by)
- PubMed Central (publishedVersion, cc-by-nc-nd)
- PubMed (publishedVersion)
- VU Research Portal (publishedVersion, other-oa)
- DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion)
- Funders
- Vrije Universiteit Amsterdam
- Topics
- Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Machine Learning in Bioinformatics
- Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Protein Structure and Dynamics
- Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Bioinformatics and Genomic Networks
- Keywords
- protein-protein interface prediction, protein language models, protein embeddings, protein-protein interaction, ablation study
- MeSH terms
- Deep Learning, Bacterial Proteins, Mycobacterium tuberculosis, Proteins, Computational Biology, Protein Interaction Mapping, Databases, Protein
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
How to cite
Cite CoChat for the checks on this page, and the paper for its own findings.
- This page
- CoChat. (2026). Citation check: PIPENN-EMB: ensemble net and protein embeddings generalise protein interface prediction beyond homology. https://app.cochat.ai/papers/10.1101/2024.10.31.621117
- The paper
- Thomas, D. P. G., Garcia Fernandez, C. M., Haydarlou, R., & Feenstra, K. A. (2024). PIPENN-EMB: ensemble net and protein embeddings generalise protein interface prediction beyond homology. Scientific Reports. https://doi.org/10.1101/2024.10.31.621117
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.