The DOI matches its Crossref record; 1 of the 1 references it cites were found in a registry or an index.
E-GuARD: Expert-Guided Augmentation for Robust Interference Compound Detection
Open accessPreprint — published as 10.1186/s13321-025-01014-3 · Paper page
Abstract
Assay interference caused by small organic compounds continues to pose formidable challenges to early drug discovery. Various computational methods have been developed to identify compounds likely to cause assay interference. However, due to the scarcity of data available for model development, the predictive accuracy and applicability of these approaches are limited. In this work, we present E-GuARD (Expert-Guided Augmentation for Robust Interference Compound Detection), a novel framework seeking to address data scarcity and imbalance by integrating self-distillation, active learning, and expert-guided molecular generation. E-GuARD iteratively enriches the training data with interference-relevant molecules, resulting in quantitative structure-interference relationship (QSIR) models with superior performance. We demonstrate the utility of E-GuARD with the examples of four high-quality data sets on thiol reactivity, redox reactivity, nanoluciferase inhibition, and firefly luciferase inhibition. Our models reached MCC values of up to 0.47 for these data sets, with two-fold or higher improvements in enrichment factors compared to models trained without E-GuARD data augmentation. These results highlight the potential of E-GuARD as a scalable solution to mitigating assay interference in early drug discovery.
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.
Integrity signals
Registry
Cites
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.
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
Flagged first, then by citations
Research works printed but not linked: 35
- Schneider G (2018) Automating drug discovery. Nat Rev Drug Discov 17:97–113 · DOI · Paper page
- Tan L, Hirte S, Palmacci V, Stork C, Kirchmair J (2024) Tackling assay interference associated with small molecules. Nat Rev Chem 8:319–339 · DOI
- Thorne N, Auld DS, Inglese J (2010) Apparent activity in high-throughput screening: origins of compound-dependent assay interference. Curr Opin Chem Biol 14:315–324 · DOI
- Baell J, Walters MA, Chemistry (2014) Chemical con artists foil drug discovery. Nature 513:481–483 · DOI
- Stork C, Mathai N, Kirchmair J (2021) Computational prediction of frequent hitters in target-based and cell-based assays. Artif Intell Life Sci 1:100007
- Stork C et al (2020) NERDD: a web portal providing access to in silico tools for drug discovery. Bioinformatics 36:1291–1292 · DOI
- Palmacci V, Hirte S, Hernández González JE, Montanari F, Kirchmair J (2024) Statistical approaches enabling technology-specific assay interference prediction from large screening data sets. Artif Intell Life Sci 5:100099
- Yang Z-Y et al (2021) ChemFLuo: a web-server for structure analysis and identification of fluorescent compounds. Brief Bioinform 22:bbaa282 · DOI
- Yang Z-Y et al (2019) Structural Analysis and identification of colloidal aggregators in drug discovery. J Chem Inf Model 59:3714–3726 · DOI
- David L et al (2019) Identification of Compounds that interfere with high-throughput screening assay technologies. ChemMedChem 14:1795–1802 · DOI
- Alves VM et al (2023) Lies and Liabilities: computational assessment of high-throughput screening hits to identify artifact compounds. J Med Chem 66:12828–12839 · DOI
- Dubey R, Zhou J, Wang Y, Thompson PM, Ye J (2014) Analysis of sampling techniques for imbalanced data: An n = 648 ADNI study. NeuroImage 87:220–241 · DOI
- Lin T-Y, Goyal P, Girshick R, He K, Dollár P (2018) Focal Loss for dense object detection. Preprint at. https://doi.org/10.48550/arXiv.1708.02002 · DOI · Paper page
- Bjerrum EJ (2017) SMILES enumeration as data augmentation for neural network modeling of molecules. Preprint at. https://doi.org/10.48550/arXiv.1703.07076 · DOI
- Schaudt D et al (2023) Augmentation strategies for an imbalanced learning problem on a novel COVID-19 severity dataset. Sci Rep 13:18299 · DOI
- Chawla NV, Bowyer KW, Hall LO, Kegelmeyer WP (2002) SMOTE: Synthetic Minority Over-sampling Technique. J Artif Intell Res 16:321–357 · DOI · Paper page
- Xie Q, Luong M-T, Hovy E, Le QV (2020) Self-training with Noisy Student improves ImageNet classification. Preprint at http://arxiv.org/abs/1911.04252 · arXiv
- Zhang L et al (2019) Be Your Own Teacher: Improve the performance of convolutional neural networks via self distillation. Preprint https://doi. org/10.48550/arXiv.1905.08094 · DOI
- Liu Y, Lim H, Xie L (2022) Exploration of chemical space with partial labeled noisy student self-training and self-supervised graph embedding. BMC Bioinformatics 23:158 · DOI
- Huang R et al (2016) Tox21Challenge to build predictive models of nuclear receptor and stress response pathways as mediated by exposure to environmental chemicals and drugs. Front Environ Sci 3 · DOI
- Fralish Z, Reker D (2024) Taking a deep dive with active learning for drug discovery. Nat Comput Sci 4:727–728 · DOI
- Nahal Y et al (2024) Human-in-the-loop active learning for goal-oriented molecule generation. Preprint at. https://doi.org/10.26434/chemrxiv-2024-623lx · DOI
- Loeffler HH et al (2024) Reinvent 4: Modern AI–driven generative molecule design. J Cheminformatics 16:20 · DOI
- Choung O-H, Vianello R, Segler M, Stiefl N (2023) Jiménez-Luna, J. Extracting medicinal chemistry intuition via preference machine learning. Nat Commun 14:6651 · DOI
- Lemaitre G, Nogueira F, Aridas CK (2016) Imbalanced-learn: a Python toolbox to tackle the curse of imbalanced datasets in machine learning. arXiv org https://arxiv.org/abs/1609.06570v1 · arXiv
- Akiba T, Sano S, Yanase T, Ohta T, Koyama M (2019) Optuna: a next-generation hyperparameter optimization framework. arXiv org https://arxiv.org/abs/1907.10902v1 · arXiv
- Bickerton GR, Paolini GV, Besnard J, Muresan S, Hopkins AL (2012) Quantifying the chemical beauty of drugs. Nat Chem 4:90–98 · DOI
- Smith FB et al (2023) Prediction-oriented bayesian active learning. Preprint at. https://doi.org/10.48550/arXiv.2304.08151 · DOI · Paper page
- Rodríguez-Pérez R, Trunzer M, Schneider N, Faller B, Gerebtzoff G (2023) Multispecies machine learning predictions of in vitro intrinsic clearance with uncertainty quantification analyses. Mol Pharm 20:383–394 · DOI
- Baell JB, Holloway GA (2010) New substructure filters for removal of Pan Assay Interference Compounds (PAINS) from screening libraries and for their exclusion in bioassays. J Med Chem 53:2719–2740 · DOI
- Kornblith S, Norouzi M, Lee H, Hinton G (2019) Similarity of neural network representations revisited. in Proceedings of the 36th International Conference on Machine Learning 3519–3529PMLR
- Welsch M, Hirte S, Kirchmair J (2024) Deciphering molecular embeddings with centered kernel alignment. J Chem Inf Model 64:7303–7312 · DOI
- Davies A, Ghahramani Z (2014) Preprint https://doi. org/10.48550/arXiv.1402.4293. The Random Forest Kernel and other kernels for big data from random partitions · DOI
- Abdullah BM, Zaitova I, Avgustinova T, Möbius B, Klakow D (2021) How familiar does that sound? Cross-lingual representational similarity analysis of acoustic word embeddings. Preprint at. https://doi.org/10.48550/arXiv.2109.10179 · DOI
- Vogt M (2023) Exploring chemical space — Generative models and their evaluation. Artif Intell Life Sci 3:100064
Metadata
- Authors
- Palmacci, V., Nahal, Y., Welsch, M., Engkvist, O., Kaski, S., & Kirchmair, J.
- Affiliations
- University of Vienna (AT)
- AstraZeneca (Sweden) (SE)
- Aalto University
- Corresponding author
- Kirchmair, Johannes
- Venue
- Journal of Cheminformatics
- Published
- 2025-01-28
- Volume
- 17
- Issue
- 1
- Pages
- 64-64
- ISSN
- 1758-2946 (ISSN-L)
- PMC
- PMC12042382
- OpenAlex
- W4409923886
- Cited by
- 6
- References
- 1
- Citations by year
- 2026: 3 · 2025: 3
- FWCI
- 3.10
- Type
- journal article
- Language
- en
- Open access
- gold
- doi.org
- Locations
- Journal of Cheminformatics (publishedVersion, cc-by)
- Research Square (acceptedVersion, cc-by)
- PubMed Central (publishedVersion, cc-by)
- PubMed (publishedVersion)
- Aaltodoc (Aalto University) (acceptedVersion, cc-by)
- Chalmers Research (Chalmers University of Technology) (submittedVersion, cc-by)
- DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
- Research Explorer (The University of Manchester) (submittedVersion, cc-by)
- Funders
- Österreichische Nationalstiftung für Forschung, Technologie und Entwicklung
- UK Research and Innovation (EP/W002973/1)
- European Commission (956832, HORIZON2020)
- Universität Wien
- Christian Doppler Forschungsgesellschaft
- Finnish Center for Artificial Intelligence
- Horizon 2020 Framework Programme (956832)
- Topics
- Life Sciences › Pharmacology, Toxicology and Pharmaceutics › Drug Discovery › Computational Drug Discovery Methods
- Physical Sciences › Materials Science › Materials Chemistry › Machine Learning in Materials Science
- Physical Sciences › Physics and Astronomy › Atomic and Molecular Physics, and Optics › Advanced Chemical Physics Studies
- Keywords
- assay interference, self-distillation, active learning (machine learning), data scarcity, class imbalance, thiol reactivity, redox reactivity, early drug discovery, data augmentation
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
