CoChat
Sign Up
Cited by 1Protein Structure and DynamicsMachine Learning in BioinformaticsComputational Drug Discovery Methods

The DOI matches its Crossref record; 6 of the 6 references it cites were found in a registry or an index.

Hybrid deep learning with protein language models and dual-path architecture for predicting IDP functions

Open access
Liang, Jiahui, Luo, Yuxian, Su, Baoquan, Peng, Zhenling·2026·Briefings in Bioinformatics
DOI 10.1093/bib/bbag126

Published version of a preprint

Open this paper in CoChat — free account

Abstract

Intrinsically disordered regions (IDRs) drive essential cellular functions but resist conventional structural-function annotation due to their dynamic conformations. Current computational methods struggle with cross-dataset generalization and functional subtype discrimination. We present IDPFunNet, a hybrid deep learning model integrating convolutional neural networks, bidirectional LSTM, residual MLP, and the protein language model ProtT5 to predict six IDR functional classes: five binding subtypes and disordered flexible linkers (DFLs). Its dual-path architecture decouples binding prediction from DFL identification. Leveraging ProtT5 evolutionary embeddings, which outperformed ESM-family models and AlphaFold2 structural features (by ≥1.3% average AUC and ≥ 12.7% average APS), IDPFunNet achieves state-of-the-art performance. Across six independent benchmarks, including CAID2/3 blind tests, it consistently surpasses existing general predictors DisoFLAG and DeepDISOBind in protein-binding prediction, with AUCs of 0.866 (TE210) and 0.832 (TE83), representing significant gains of 1.5%–8.1% in AUC and 13.5%–26.7% in APS (p-value < 0.05), while remaining competitive with specialized DFL predictors. Further analyses show multi-task learning enhances protein/lipid/small molecule-binding (3.1%–35.1% AUC gains), BiLSTMs are optimal for DFL identification, and self-attention shows potential for nucleic acid-binding (AUC 0.831). IDPFunNet thus provides an interpretable and generalizable framework for comprehensive IDR functional mapping. The webserver of IDPFunNet is freely available at https://yanglab.qd.sdu.edu.cn/IDPFunNet/ and the standalone package can be downloaded from https://github.com/IDRIDP/IDPFunNet/tree/master.

Overview

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1093/bib/bbag126

Cites

6 of 6 references verified0 citation errors · 0 not checkable

Statistics

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

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

Flagged first, then by citations

Show all 6 references

Research works printed but not linked: 48

  1. Oldfield, 2014, Intrinsically disordered proteins and intrinsically disordered protein regions, Annu Rev Biochem, 83, 553, 10.1146/annurev-biochem-072711-164947 · DOI · Paper page
  2. Uversky, 2014, Introduction to intrinsically disordered proteins (IDPs), Chem Rev, 114, 6557, 10.1021/cr500288y · DOI
  3. Lee, 2014, Classification of intrinsically disordered regions and proteins, Chem Rev, 114, 6589, 10.1021/cr400525m · DOI
  4. Wright, 2015, Intrinsically disordered proteins in cellular signalling and regulation, Nat Rev Mol Cell Biol, 16, 18, 10.1038/nrm3920 · DOI
  5. Babu, 2016, The contribution of intrinsically disordered regions to protein function, cellular complexity, and human disease, Biochem Soc Trans, 44, 1185, 10.1042/BST20160172 · DOI
  6. Holehouse, 2024, The molecular basis for cellular function of intrinsically disordered protein regions, Nat Rev Mol Cell Biol, 25, 187, 10.1038/s41580-023-00673-0 · DOI
  7. Fung, 2018, IDPs in macromolecular complexes: the roles of multivalent interactions in diverse assemblies, Curr Opin Struct Biol, 49, 36, 10.1016/j.sbi.2017.12.007 · DOI
  8. Orand, 2025, Binding mechanisms of intrinsically disordered proteins: insights from experimental studies and structural predictions, Curr Opin Struct Biol, 90, 10.1016/j.sbi.2024.102958 · DOI
  9. Uversky, 2018, Intrinsic disorder, protein-protein interactions, and disease, Adv Protein Chem Struct Biol, 110, 85, 10.1016/bs.apcsb.2017.06.005 · DOI
  10. Raj, 2017, The transactivation domains of the p53 protein, Cold Spring Harb Perspect Med, 7, 10.1101/cshperspect.a026047 · DOI
  11. Hilser, 2007, Intrinsic disorder as a mechanism to optimize allosteric coupling in proteins, Proc Natl Acad Sci U S A, 104, 8311, 10.1073/pnas.0700329104 · DOI · Paper page
  12. Ferreon, 2013, Modulation of allostery by protein intrinsic disorder, Nature, 498, 390, 10.1038/nature12294 · DOI
  13. Terakawa, 2014, Multi-scale ensemble Modeling of modular proteins with intrinsically disordered linker regions: application to p53, Biophys J, 107, 721, 10.1016/j.bpj.2014.06.026 · DOI · Paper page
  14. Wang, 2011, Intrinsic disorder mediates the diverse regulatory functions of the Cdk inhibitor p21, Nat Chem Biol, 7, 214, 10.1038/nchembio.536 · DOI
  15. Mészáros, 2009, Prediction of protein binding regions in disordered proteins, PLoS Comput Biol, 5, 5, 10.1371/journal.pcbi.1000376 · DOI
  16. Peng, 2015, High-throughput prediction of RNA, DNA and protein binding regions mediated by intrinsic disorder, Nucleic Acids Res, 43, e121, 10.1093/nar/gkv585 · DOI
  17. Malhis, 2016, MoRFchibi SYSTEM: software tools for the identification of MoRFs in protein sequences, Nucleic Acids Res, 44, W488, 10.1093/nar/gkw409 · DOI
  18. Meng, 2016, DFLpred: high-throughput prediction of disordered flexible linker regions in protein sequences, Bioinformatics, 32, i341, 10.1093/bioinformatics/btw280 · DOI
  19. Peng, 2020, APOD: accurate sequence-based predictor of disordered flexible linkers, Bioinformatics, 36, i754
  20. Perovic, 2018, IDPpi: protein-protein interaction analyses of human intrinsically disordered proteins, Sci Rep, 8, 8, 10.1038/s41598-018-28815-x · DOI
  21. Peng, 2023, CLIP: accurate prediction of disordered linear interacting peptides from protein sequences using co-evolutionary information, Brief Bioinform, 24, 10.1093/bib/bbac502 · DOI
  22. Hu, 2021, flDPnn: accurate intrinsic disorder prediction with putative propensities of disorder functions, Nat Commun, 12, 4438, 10.1038/s41467-021-24773-7 · DOI
  23. Wang, 2024, flDPnn2: accurate and fast predictor of intrinsic disorder in proteins, J Mol Biol, 436, 10.1016/j.jmb.2024.168605 · DOI
  24. Zhang, 2022, DeepDISOBind: accurate prediction of RNA-, DNA- and protein-binding intrinsically disordered residues with deep multi-task learning, Brief Bioinform, 23, 23, 10.1093/bib/bbab521 · DOI
  25. Katuwawala, 2021, DisoLipPred: accurate prediction of disordered lipid-binding residues in protein sequences with deep recurrent networks and transfer learning, Bioinformatics, 38, 115, 10.1093/bioinformatics/btab640 · DOI
  26. Mehdiabadi, 2026, Critical assessment of protein intrinsic disorder round 3-predicting disorder in the era of protein language models, Proteins, 94, 414, 10.1002/prot.70045 · DOI
  27. Sharma, 2023, DRPBind: prediction of DNA, RNA and protein binding residues in intrinsically disordered protein sequences, 10.1101/2023.03.20.533427 · DOI
  28. Piovesan, 2022, Intrinsic protein disorder and conditional folding in AlphaFoldDB, Protein Sci, 31, e4466, 10.1002/pro.4466 · DOI · Paper page
  29. Hanson, 2019, SPOT-Disorder2: improved protein intrinsic disorder prediction by Ensembled deep learning, Genomics Proteomics Bioinformatics, 17, 645, 10.1016/j.gpb.2019.01.004 · DOI · Paper page
  30. Rives, 2021, Biological structure and function emerge from scaling unsupervised learning to 250 million protein sequences, Proc Natl Acad Sci U S A, 118, 10.1073/pnas.2016239118 · DOI · Paper page
  31. Elnaggar, 2022, ProtTrans: toward understanding the language of life through self-supervised learning, IEEE Trans Pattern Anal Mach Intell, 44, 7112, 10.1109/TPAMI.2021.3095381 · DOI · Paper page
  32. Xu, 2023, Accurate and fast prediction of intrinsically disordered protein by multiple protein language models and ensemble learning, J Chem Inform Model, 64, 2901, 10.1021/acs.jcim.3c01202 · DOI
  33. Pang, 2024, DisoFLAG: accurate prediction of protein intrinsic disorder and its functions using graph-based interaction protein language model, BMC Biol, 22, 3, 10.1186/s12915-023-01803-y · DOI
  34. Chakrabarti, 2022, Intrinsically disordered proteins/regions and insight into their biomolecular interactions, Biophys Chem, 283, 106769, 10.1016/j.bpc.2022.106769 · DOI
  35. Dunker, 2002, Intrinsic disorder and protein function, Biochemistry, 41, 6573, 10.1021/bi012159+ · DOI · Paper page
  36. Fatafta, 2021, Molecular simulations of IDPs: from ensemble generation to IDP interactions leading to disorder-to-order transitions, Prog Mol Biol Trans Sci, 183, 135, 10.1016/bs.pmbts.2021.06.003 · DOI
  37. Guo, 2022, AlphaFold2 models indicate that protein sequence determines both structure and dynamics, Sci Rep, 12, 10696, 10.1038/s41598-022-14382-9 · DOI · Paper page
  38. Advances in Neural Information Processing Systems 30 (NIPS 2017)
  39. Aspromonte, 2024, DisProt in 2024: improving function annotation of intrinsically disordered proteins, Nucleic Acids Res, 52, D434, 10.1093/nar/gkad928 · DOI
  40. Li, 2006, Cd-hit: a fast program for clustering and comparing large sets of protein or nucleotide sequences, Bioinformatics, 22, 1658, 10.1093/bioinformatics/btl158 · DOI
  41. Kabir, 2026, ESMDisPred: a structure-aware CNN-transformer architecture for intrinsically disordered protein prediction, 10.64898/2026.01.22.701204 · DOI
  42. Kulmanov, 2020, DeepGOPlus: improved protein function prediction from sequence, Bioinformatics, 36, 422, 10.1093/bioinformatics/btz595 · DOI
  43. Dhanuka, 2023, A comprehensive survey of deep learning techniques in protein function prediction, IEEE/ACM Trans Comput Biol Bioinform, 20, 2291, 10.1109/TCBB.2023.3247634 · DOI
  44. Zhu, 2022, Integrating unsupervised language model with triplet neural networks for protein gene ontology prediction, PLoS Comput Biol, 18, 18, 10.1371/journal.pcbi.1010793 · DOI
  45. Ke, 2024, Prediction of protein N-terminal acetylation modification sites based on CNN-BiLSTM-attention model, Comput Biol Med, 174, 108330, 10.1016/j.compbiomed.2024.108330 · DOI
  46. Fan, 2024, Predicting the functional changes in protein mutations through the application of BiLSTM and the self-attention mechanism, Ann Data Sci, 11, 1077, 10.1007/s40745-024-00530-7 · DOI
  47. Ilzhofer, 2022, SETH predicts nuances of residue disorder from protein embeddings, Front Bioinform, 2, 1019597, 10.3389/fbinf.2022.1019597 · DOI
  48. Wang, 2024, Assessment of disordered linker predictions in the CAID2 experiment, Biomolecules, 15, 14, 10.3390/biom15010014 · DOI

Metadata

Authors
Liang, J., Luo, Y., Su, B., & Peng, Z.
Affiliations
Shandong University (CN)
Corresponding author
Peng, Zhenling
ORCID
Venue
Briefings in Bioinformatics
Publisher
Oxford University Press (OUP)
Published
2026-03-01
Volume
27
Issue
2
ISSN
  • 1467-5463 (ISSN-L)
  • 1477-4054
DOI
10.1093/bib/bbag126
OpenAlex
W7147485138
Cited by
1
References
6
Citations by year
2026: 1
FWCI
2.07
Type
journal article
Language
en
Open access
Licence
Locations
  • Briefings in Bioinformatics (publishedVersion, cc-by)
  • bioRxiv (Cold Spring Harbor Laboratory) (acceptedVersion, cc-by-nc-nd)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
Funders
  • National Key Research and Development Program of China (2023YFF1204003)
  • National Natural Science Foundation of China (T2222012)
  • Fundamental Research Funds for the Central Universities
Topics
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Protein Structure and Dynamics
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Machine Learning in Bioinformatics
  • Life Sciences › Pharmacology, Toxicology and Pharmaceutics › Drug Discovery › Computational Drug Discovery Methods
Keywords
intrinsically disordered regions, protein function prediction, protein language models, ProtT5, protein binding prediction, dual-path architecture, convolutional neural networks, bidirectional LSTM, multi-task learning, cross-dataset generalization
MeSH terms
Deep Learning, Convolutional Neural Networks, Long Short Term Memory, Humans, Protein Binding, Software, Neural Networks, Computer, Computational Biology, Intrinsically Disordered Proteins

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