Hybrid Network Model for “Deep Learning” of Chemical Data: Application to Antimicrobial Peptides
The DOI matches its Crossref record; 58 of the 59 references it cites were found in a registry or an index.
Abstract
We present a “deep” network architecture for chemical data analysis and classification together with a prospective proof‐of‐concept application. The model features a self‐organizing map (SOM) as the input layer of a feedforward neural network. The SOM converts molecular descriptors to a two‐dimensional image for further processing. We implemented lateral neuron inhibition for contrast enhancement. The model achieved improved classification accuracy and predictive robustness compared to feedforward network classifiers lacking the SOM layer. By nonlinear dimensionality reduction the networks extracted meaningful chemical features from the data and outperformed linear principal component analysis (PCA). The learning machine was trained on the sequence‐length independent recognition of antibacterial peptides and correctly predicted the killing activity of a synthetic test peptide against Staphylococcus aureus in an in vitro experiment.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
In the citing paper
“In previous studies, it has been verified that LSTM-based neural networks could learn AMP’s amino-acid grammar and create new potential AMP sequences32,33.”
From Experimental demonstration of highly reliable dynamic memristor for artificial neuron and neuromorphic computing (2022) · cites this paper as [33]
Citations
Cites
LeCun, Y., Bengio, Y., & Hinton, G. (2015) · Nature · doi:10.1038/nature14539 · Cited by 84,126
Human-level control through deep reinforcement learning
Mnih, V., Kavukcuoglu, K., Silver, D., Rusu, A. A., Veness, J., Bellemare, M. G., Graves, A., Riedmiller, M., Fidjeland, A. K., Ostrovski, G., et al. (2015) · Nature · doi:10.1038/nature14236 · Cited by 30,357
Reducing the dimensionality of data with neural networks
Hinton, G. E., & Salakhutdinov, R. R. (2006) · Science · doi:10.1126/science.1127647 · Cited by 20,928
Wold, S., Esbensen, K., & Geladi, P. (1987) · Chemometrics and Intelligent Laboratory Systems · doi:10.1016/0169-7439(87)80084-9 · Cited by 11,606
Self-organized formation of topologically correct feature maps
Kohonen, T. (1982) · Biological Cybernetics · doi:10.1007/bf00337288 · Cited by 9,529
Learning Deep Architectures for AI
Bengio, Y. (2009) · Foundations and Trends® in Machine Learning · doi:10.1561/2200000006 · Cited by 9,239
Comparison of the predicted and observed secondary structure of T4 phage lysozyme
Matthews, B. W. (1975) · Biochimica et Biophysica Acta (BBA) - Protein Structure · doi:10.1016/0005-2795(75)90109-9 · Cited by 5,399
Using circular dichroism spectra to estimate protein secondary structure
Greenfield, N. J. (2006) · Nature Protocols · doi:10.1038/nprot.2006.202 · Cited by 4,014
Self-organized formation of topographically correct feature maps
Kohonen, T. (1982) · Biological Cybernetics · Cited by 2,995
Evolution strategies – A comprehensive introduction
Beyer, H.-G., & Schwefel, H.-P. (2002) · Natural Computing · doi:10.1023/a:1015059928466 · Cited by 2,790
Research works printed but not linked: 3
- Hecht-Nielsen R., 1987, Proceedings of the IEEE First International Conference on Neural Networks, 19
- Kernighan B. W., 1988, The C Programming Language
- Schneider G., 1999, From Symbiosis to Eukaryotism – Endocytobiology, 589
Cited by
Generating Focused Molecule Libraries for Drug Discovery with Recurrent Neural Networks
Segler, M. H. S., Kogej, T., Tyrchan, C., & Waller, M. P. (2018) · ACS Central Science · doi:10.1021/acscentsci.7b00512 · Cited by 1,537
Peptide Design Principles for Antimicrobial Applications
Torres, M. D. T., Sothiselvam, S., Lu, T. K., & de la Fuente-Nunez, C. (2019) · Journal of Molecular Biology · doi:10.1016/j.jmb.2018.12.015 · Cited by 438
Park, S.-O., Jeong, H., Park, J., Bae, J., & Choi, S. (2022) · Nature Communications · doi:10.1038/s41467-022-30539-6 · Cited by 331
“In previous studies, it has been veri fi ed that LSTM-based neural networks could learn AMP ’ s amino-acid grammar and create new potential AMP sequences 32,33 .”
A renaissance of neural networks in drug discovery
Baskin, I. I., Winkler, D., & Tetko, I. V. (2016) · Expert Opinion on Drug Discovery · doi:10.1080/17460441.2016.1201262 · Cited by 235
“At the present time, however, there is a clear trend towards their convergence [80].”
BackgroundMachine learning designs non-hemolytic antimicrobial peptides
Capecchi, A., Cai, X., Personne, H., Köhler, T., van Delden, C., & Reymond, J.-L. (2021) · Chemical Science · doi:10.1039/d1sc01713f · Cited by 173
Machine learning and data science in soft materials engineering
Ferguson, A. L. (2018) · Journal of Physics: Condensed Matter · doi:10.1088/1361-648x/aa98bd · Cited by 173
A review on antimicrobial peptides databases and the computational tools
Ramazi, S., Mohammadi, N., Allahverdi, A., Khalili, E., & Abdolmaleki, P. (2022) · Database · doi:10.1093/database/baac011 · Cited by 176
“Together, these studies highlight a diversity of methods and approaches that have been used to classify and design AMPs with great success (128).”
Concepts and applications of chemical fingerprint for hit and lead screening
Yang, J., Cai, Y., Zhao, K., Xie, H., & Chen, X. (2022) · Drug Discovery Today · doi:10.1016/j.drudis.2022.103356 · Cited by 167
Lee, E. Y., Lee, M. W., Fulan, B. M., Ferguson, A. L., & Wong, G. C. L. (2017) · Interface Focus · doi:10.1098/rsfs.2016.0153 · Cited by 156
Synthetic Biology and Computer-Based Frameworks for Antimicrobial Peptide Discovery
Torres, M. D. T., Cao, J., Franco, O. L., Lu, T. K., & de la Fuente‐Núñez, C. (2021) · ACS Nano · doi:10.1021/acsnano.0c09509 · Cited by 126
Fernández, M., Ban, F., Woo, G., Hsing, M., Yamazaki, T., Leblanc, E., Rennie, P. S., Welch, W. J., & Cherkasov, A. R. (2018) · Journal of Chemical Information and Modeling · doi:10.1021/acs.jcim.8b00338 · Cited by 125
Variational Autoencoder for Generation of Antimicrobial Peptides
Dean, S. N., & Walper, S. A. (2020) · ACS Omega · doi:10.1021/acsomega.0c00442 · Cited by 121
Deep Learning for Novel Antimicrobial Peptide Design
Wang, C. E., Garlick, S., & Zloh, M. (2021) · Biomolecules · doi:10.3390/biom11030471 · Cited by 128
Machine Learning Prediction of Antimicrobial Peptides
Wang, G., Vaisman, I. I., & van Hoek, M. L. (2022) · Methods in molecular biology · doi:10.1007/978-1-0716-1855-4_1 · Cited by 131
Encodings and models for antimicrobial peptide classification for multi-resistant pathogens
Spänig, S., & Heider, D. (2019) · BioData Mining · doi:10.1186/s13040-019-0196-x · Cited by 107
“Afterwards, the continuous SOM responses are adjusted by means of lateral inhibition and utilized as input for a deep learning model in order to predict helical AMPs [85].”
BackgroundMethodInfluentialMachine learning-enabled discovery and design of membrane-active peptides
Lee, E. Y., Wong, G. C. L., & Ferguson, A. L. (2017) · Bioorganic & Medicinal Chemistry · doi:10.1016/j.bmc.2017.07.012 · Cited by 100
“reported a two-step process wherein a self-organizing map (SOM) was used to perform nonlinear dimensionality reduction over 147 peptide descriptors as a preprocessing step prior to ANN classification.(34) In our own recent work, we developed a SVM classifier to predict a-helical AMP activity with 90% accuracy.”
MethodToward computer-made artificial antibiotics
Torres, M. D. T., & de la Fuente‐Núñez, C. (2019) · Current Opinion in Microbiology · doi:10.1016/j.mib.2019.03.004 · Cited by 89
Wang, B., Lin, P., Zhong, Y., Tan, X., Shen, Y., Huang, Y., Jin, K., Zhang, Y., Zhan, Y., Shen, D., et al. (2025) · Nature Microbiology · doi:10.1038/s41564-024-01907-3 · Cited by 96
“These models incorporate techniques such as Long Short-Term Memory (LSTM), Recurrent Neural Network (RNN), Attention, Bidirectional Encoder Representation from Transformers (BERT) and other custom neural networks 25,27–32 .”
Deep Learning-Based Bioactive Therapeutic Peptide Generation and Screening
Zhang, H., Saravanan, K. M., Wei, Y., Jiao, Y., Yang, Y., Pan, Y. F., Wu, X., & Zhang, J. Z. H. (2023) · Journal of Chemical Information and Modeling · doi:10.1021/acs.jcim.2c01485 · Cited by 74
Lin, Y., Cai, Y.-Y., Liu, J., Lin, C., & Liu, X. (2019) · BMC Bioinformatics · doi:10.1186/s12859-019-2766-9 · Cited by 70
Deep learning for mining protein data
Shi, Q., Chen, W., Huang, S., Wang, Y., & Xue, Z. (2019) · Briefings in Bioinformatics · doi:10.1093/bib/bbz156 · Cited by 72
Rodríguez‐González, A., Zanin, M., & Menasalvas, E. (2019) · Yearbook of Medical Informatics · doi:10.1055/s-0039-1677910 · Cited by 75
“Several artificial intelligence approaches have been tried in the last years, including artificial neural networks and their LSTM variant [43, 44], evolutionary algorithms [45], and genetic programming [46].”
BackgroundUnnatural amino acids: promising implications for the development of new antimicrobial peptides
Wang, X., Yang, X., Wang, Q., & Meng, D. (2022) · Critical Reviews in Microbiology · doi:10.1080/1040841x.2022.2047008 · Cited by 52
Introducing self-organized maps (SOM) as a visualization tool for materials research and education
Qian, J., Nguyen, N. P., Oya, Y., Kikugawa, G., Okabe, T., Huang, Y., & Ohuchi, F. S. (2019) · Results in Materials · doi:10.1016/j.rinma.2019.100020 · Cited by 58
Dias, T., Gaudêncio, S. P., & Pereira, F. (2018) · Marine Drugs · doi:10.3390/md17010016 · Cited by 43
“, indolylidinepyrazolone derivatives [12], N2-acyl isonicotinic acid hydrazide derivatives [13], (Z)-2-(nitroimidazolylmethylene)3(2H)-benzofuranone derivatives [14], peptides [15,16], anthraquinone derivative [17], indole carboxamide 4-piperazine derivatives [18].”
BackgroundPuentes, P. R., Henao, M. C., Torres, C. E., Gómez, S. C., Gómez, L. A., Burgos, J. C., Arbeláez, P., Osma, J. F., Muñoz-Camargo, C., Reyes, L. H., et al. (2020) · Antibiotics · doi:10.3390/antibiotics9120854 · Cited by 42
“Specifically, for AMP discovery, peptides have been represented by their topological pharmacophore features, converting a given residue sequence to a 147-dimensional descriptor that encodes the standardized cross-correlated pharmacophore features of the amino acid residues [121].”
BackgroundVirtual Screening Meets Deep Learning
Pérez-Sianes, J., Pérez‐Sánchez, H., & Díaz, F. (2018) · Current Computer - Aided Drug Design · doi:10.2174/1573409914666181018141602 · Cited by 40
Macromolecular target prediction by self-organizing feature maps
Schneider, G., & Schneider, P. (2016) · Expert Opinion on Drug Discovery · doi:10.1080/17460441.2017.1274727 · Cited by 34
“In the first case, the SOM response pattern serves as a new molecular descriptor as input to a target prediction model [43].”
BackgroundSelf-citationSöylemez, Ü. G., Yousef, M., Kesmen, Z., Büyükkiraz, M. E., & Bakır-Güngör, B. (2022) · Applied Sciences · doi:10.3390/app12073631 · Cited by 35
Integrated computational approaches for advancing antimicrobial peptide development
Fang, Y., Ma, Y., Yu, K., Dong, J. N., & Zeng, W. (2024) · Trends in Pharmacological Sciences · doi:10.1016/j.tips.2024.09.011 · Cited by 27
Lee, M. W., Lee, E. Y., Ferguson, A. L., & Wong, G. C. L. (2018) · Current Opinion in Colloid & Interface Science · doi:10.1016/j.cocis.2018.11.003 · Cited by 22
Tuning gradient boosting for imbalanced bioassay modelling with custom loss functions
Boldini, D., Friedrich, L., Kühn, D., & Sieber, S. A. (2022) · Journal of Cheminformatics · doi:10.1186/s13321-022-00657-w · Cited by 21
“Thanks to these resources, it is straightforward to obtain thousands of training points to develop high-performing predictive models, which can then be used to screen for novel ligands, antibiotics, antivirals and so forth [6–8].”
BackgroundStructure–activity relationships and drug design
Gianti, E., & Zauhar, R. J. (2020) · Elsevier eBooks · doi:10.1016/b978-0-12-820007-0.00007-6 · Cited by 19
Fabrizio, A., Meyer, B., & Corminbœuf, C. (2020) · The Journal of Chemical Physics · doi:10.1063/5.0005039 · Cited by 13
Molecular activity prediction using deep learning software library
Kato, Y., Hamada, S., & Goto, M. (2016) · International Conference on Advanced Informatics: Concepts, Theory and Applications (ICAICTA) · doi:10.1109/icaicta.2016.7803113 · Cited by 10
“Appearances of general purpose deep learning software libraries, TensorFlowTM [2],Chainer [3],Caffe [4], are also expected to further development for more practical usage.”
BackgroundGEU-AMP50: Enhanced antimicrobial peptide prediction using a machine learning approach
Panwar, S., Thapliyal, M., Kuriyal, V., Tripathi, V., & Thapliyal, A. (2022) · Materials Today Proceedings · doi:10.1016/j.matpr.2022.09.326 · Cited by 9
Carrera, G. V. S. M., & da Ponte, M. N. (2020) · Chemistry - Methods · doi:10.1002/cmtd.202000031 · Cited by 9
Pandey, P., & Srivastava, A. K. (2024) · Proteins Structure Function and Bioinformatics · doi:10.1002/prot.26681 · Cited by 16
“34 In 2017, Meher et al. developed a more efficient SVM-based AMP prediction model using PseAAC feature vector by incorporating compositional, physicochemical, and structural features.”
Amin, A. E. (2020) · International journal of intelligent computing and information sciences/International Journal of Intelligent Computing and Information Sciences · doi:10.21608/ijicis.2020.23149.1015 · Cited by 6
“The architecture of SOM consists of single layer feedforward networks while the output layer that is usually 2D or 3D are arranged in a low dimensional grid [10].”
BackgroundMethacrylate Coatings for Titanium Surfaces to Optimize Biocompatibility
Sun, A., Ashammakhi, N. A., & Dokmeci, M. R. (2020) · Micromachines · doi:10.3390/mi11010087 · Cited by 6
Machine learning for synthetic biology: Methods and applications
Hu, R., Zhang, S., Meng, H., Han, Y., Zhang, J., Luo, X., Si, T., Liu, C., & Qiao, Y. (2020) · Chinese Science Bulletin (Chinese Version) · doi:10.1360/tb-2020-0456 · Cited by 7
Identification of Synthetic Activators of Cancer Cell Migration by Hybrid Deep Learning
Bruns, D., Gawehn, E., Kumar, K. S., Schneider, P., Baumgartner, M., & Schneider, G. (2019) · ChemBioChem · doi:10.1002/cbic.201900346 · Cited by 5
Self-citationPandey, P., & Srivastava, A. K. (2023) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2023.06.04.543607 · Cited by 4
“In 2016, Schneider et al. proposed self-organizing self-organizing maps (SOM) based deep network to analyze AMPs[26].”
BackgroundAmpClass: an Antimicrobial Peptide Predictor Based on Supervised Machine Learning
Mera, C., Ordúz, S., Cardona, P., Orrego, A., Muñoz‐Pérez, J. M., & Branch, J. W. (2024) · Anais da Academia Brasileira de Ciências · doi:10.1590/0001-3765202420230756 · Cited by 6
Khorramfard, A., Pirgazi, J., Sorkhi, A. G., & Pourseif, M. M. (2026) · Computational Biology and Chemistry · doi:10.1016/j.compbiolchem.2026.109129 · Cited by 2
[Retracted] Impact of ANN in Revealing of Viral Peptides
Rajkumar, M., Bhukya, S. N., Ahalya, N., Elumalai, G., Sivanandam, K., Almutairi, K. M., Alonazi, W. B., Soma, S. R., & Urugo, M. M. (2022) · BioMed Research International · doi:10.1155/2022/7760734 · Cited by 2
“They [23] used unsupervised – supervised two-step models for the fi rst time to categorise AMPs, which was a fi rst in the fi eld.”
RetractedJiang, L., Yang, S., Li, Y., Zhang, Y., Ren, X., Liu, W. T., Fu, H., Bao, C., Qi, Q., Liu, T., et al. (2026) · Frontiers in Microbiology · doi:10.3389/fmicb.2026.1861239 · Cited by 2
“In 2017, Petra Schneider and colleagues introduced deep learning (DL) frameworks into AMP research through the application of deep neural network architectures for peptide classification and chemical feature analysis (Schneider et al., 2017).”
Machine Learning Designs Non-Hemolytic Antimicrobial PeptidesPreprint
Capecchi, A., Cai, X., Personne, H., Köhler, T., van Delden, C., & Reymond, J. (2021) · ChemRxiv · doi:10.26434/chemrxiv.14233418 · Cited by 1
Boldini, D., Friedrich, L., Kuhn, D., & Sieber, S. A. (2022) · Figshare · doi:10.6084/m9.figshare.21541836
Liu, S., Gao, X., Wang, J., Li, Z., Liu, Y., Wang, Y., Hou, T., Liu, Y., & Liu, Y. (2026) · Journal of Chemical Information and Modeling · doi:10.1021/acs.jcim.6c01300
Computational Platforms for Membrane-Active Peptide Therapeutic Design
Zan, B., Chen, C. H., Sabha, M., Wu, K., Kadri, M., Nwajiaku, O. L., Hoogland, D., Ulmschneider, J. P., & Ulmschneider, M. B. (2026) · Chemical Reviews · doi:10.1021/acs.chemrev.6c00100
Metadata
- Authors
- Schneider, P., Müller, A. T., Gabernet, G., Button, A. L., Posselt, G., Wessler, S., Hiss, J. A., & Schneider, G.
- Affiliations
- ETH Zurich (CH)
- University of Salzburg (AT)
- Corresponding author
- Schneider, Gisbert
- ORCID
- Venue
- Molecular Informatics
- Publisher
- Wiley
- Published
- 2016-03-10
- Volume
- 36
- Issue
- 1-2
- ISSN
- 1868-1743 (ISSN-L)
- 1868-1751
- OpenAlex
- W2302124239
- Cited by
- 52
- References
- 59
- Citations by year
- 2026: 4 · 2025: 1 · 2024: 3 · 2023: 2 · 2022: 10 · 2021: 4 · 2020: 9 · 2019: 8 · 2018: 4 · 2017: 4 · 2016: 3
- FWCI
- 5.83
- Type
- journal article
- Language
- en
- Open access
- closed
- Locations
- Molecular Informatics (publishedVersion)
- PubMed (publishedVersion)
- Funders
- Swiss National Science Foundation (200021_157190, CRSII2_160699)
- Topics
- Life Sciences › Pharmacology, Toxicology and Pharmaceutics › Drug Discovery › Computational Drug Discovery Methods
- Life Sciences › Immunology and Microbiology › Microbiology › Antimicrobial Peptides and Activities
- Physical Sciences › Computer Science › Artificial Intelligence › Neural Networks and Applications
- Keywords
- hybrid network architecture, self-organizing maps, feedforward neural networks, molecular descriptors, nonlinear dimensionality reduction, antibacterial peptides, antimicrobial peptide prediction, Staphylococcus aureus, principal component analysis comparison
- MeSH terms
- Machine Learning, Staphylococcus aureus, Antimicrobial Cationic Peptides, Principal Component Analysis
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
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.