Explainable artificial intelligence for genomic prediction of antimicrobial resistance in Escherichia coli: a leakage-controlled, cross-model, and biologically validated pilot framework

Mesut Ersin Sönmez

Explainable artificial intelligence for genomic prediction of antimicrobial resistance in Escherichia coli: a leakage-controlled, cross-model, and biologically validated pilot framework

Abstract

Antimicrobial resistance threatens modern medicine, and Escherichia coli is a leading resistant pathogen for which genome based prediction can complement slow culture testing. To be clinically useful, a predictor must be accurate, explainable, biologically coherent, and free from the optimistic bias caused by bacterial population structure. This pilot study presents a transparent framework that combines leakage controlled modelling with multi method explainability and biological validation, and is intended as a methodological guide rather than a new laboratory finding. Using laboratory confirmed susceptibility phenotypes for 58,750 to 76,364 E. coli genomes obtained from the BV-BRC database, resistance to five antibiotics was predicted from a resistance gene presence and absence matrix. Eight classifiers were tuned with Optuna under clone aware cross validation, evaluated with the Matthews correlation coefficient and complementary metrics, and interpreted with SHAP and permutation importance. Attribution stability was quantified by the cross model Jaccard overlap, and the most important genes were tested for resistance mechanism and KEGG/BRITE pathway over representation. Tree ensembles performed best, reaching a Matthews correlation coefficient of 0.94 for gentamicin and 0.92 for ampicillin, whereas ciprofloxacin remained the hardest task despite a deceptively high ROC area (area under the receiver operating characteristic curve of 0.89 versus a Matthews correlation coefficient of 0.58). Attributions were stable across architecturally different models and concentrated on the established determinants of each drug, which mechanism and pathway enrichment confirmed. The ciprofloxacin case shows that, when resistance is driven by point mutations, the data representation can matter more than the model, an interpretive lesson central to deploying artificial intelligence for resistance prediction.

Keywords

References

[1] C. J. L. Murray et al., “Global burden of bacterial antimicrobial resistance in 2019: a systematic analysis,” The Lancet, vol. 399, no. 10325, pp. 629–655, 2022, doi: 10.1016/S0140-6736(21)02724-0.
[2] H. Sati et al., “The WHO Bacterial Priority Pathogens List 2024: a prioritisation study to guide research, development, and public health strategies against antimicrobial resistance,” The Lancet Infectious Diseases, vol. 25, no. 9, pp. 1033–1043, 2025, doi: 10.1016/S1473-3099(25)00118-5.
[3] R. D. Olson et al., “Introducing the Bacterial and Viral Bioinformatics Resource Center (BV-BRC): a resource combining PATRIC, IRD and ViPR,” Nucleic Acids Research, vol. 51, no. D1, pp. D678–D689, 2023, doi: 10.1093/nar/gkac1003.
[4] M. N. Anahtar, J. H. Yang, and S. Kanjilal, “Applications of machine learning to the problem of antimicrobial resistance: an emerging model for translational research,” Journal of Clinical Microbiology, vol. 59, no. 7, p. e01260-20, 2021, doi: 10.1128/JCM.01260-20.
[5] C. M. Ardila, P. K. Yadalam, and D. González-Arroyave, “Integrating whole genome sequencing and machine learning for predicting antimicrobial resistance in critical pathogens: a systematic review of antimicrobial susceptibility tests,” PeerJ, vol. 12, p. e18213, 2024, doi: 10.7717/peerj.18213.
[6] D. Moradigaravand, M. Palm, A. Farewell, V. Mustonen, J. Warringer, and L. Parts, “Prediction of antibiotic resistance in Escherichia coli from large-scale pan-genome data,” PLoS Computational Biology, vol. 14, no. 12, p. e1006258, 2018, doi: 10.1371/journal.pcbi.1006258.
[7] P.-J. Van Camp, D. B. Haslam, and A. Porollo, “Prediction of antimicrobial resistance in Gram-negative bacteria from whole-genome sequencing data,” Frontiers in Microbiology, vol. 11, p. 1013, 2020, doi: 10.3389/fmicb.2020.01013.
[8] Y. Ren et al., “Prediction of antimicrobial resistance based on whole-genome sequencing and machine learning,” Bioinformatics, vol. 38, no. 2, pp. 325–334, 2022, doi: 10.1093/bioinformatics/btab681.

[9] D. Aytan-Aktug, P. T. L. C. Clausen, V. Bortolaia, F. M. Aarestrup, and O. Lund, “Prediction of acquired antimicrobial resistance for multiple bacterial species using neural networks,” mSystems, vol. 5, no. 1, p. e00774-19, 2020, doi: 10.1128/mSystems.00774-19.
[10] B. P. Alcock et al., “CARD 2023: expanded curation, support for machine learning, and resistome prediction at the Comprehensive Antibiotic Resistance Database,” Nucleic Acids Research, vol. 51, no. D1, pp. D690–D699, 2023, doi: 10.1093/nar/gkac920.
[11] M. C. J. Maiden et al., “Multilocus sequence typing: a portable approach to the identification of clones within populations of pathogenic microorganisms,” Proceedings of the National Academy of Sciences, vol. 95, no. 6, pp. 3140–3145, 1998, doi: 10.1073/pnas.95.6.3140.
[12] L. Breiman, “Random forests,” Machine Learning, vol. 45, no. 1, pp. 5–32, 2001, doi: 10.1023/A:1010933404324.
[13] H. Peng, F. Long, and C. Ding, “Feature selection based on mutual information: criteria of max-dependency, max-relevance, and min-redundancy,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 27, no. 8, pp. 1226–1238, 2005, doi: 10.1109/TPAMI.2005.159.
[14] S. M. Lundberg et al., “From local explanations to global understanding with explainable AI for trees,” Nature Machine Intelligence, vol. 2, no. 1, pp. 56–67, 2020, doi: 10.1038/s42256-019-0138-9.
[15] D. Chicco and G. Jurman, “The advantages of the Matthews correlation coefficient (MCC) over F1 score and accuracy in binary classification evaluation,” BMC Genomics, vol. 21, p. 6, 2020, doi: 10.1186/s12864-019-6413-7.
[16] B. W. Matthews, “Comparison of the predicted and observed secondary structure of T4 phage lysozyme,” Biochimica et Biophysica Acta (BBA) - Protein Structure, vol. 405, no. 2, pp. 442–451, 1975, doi: 10.1016/0005-2795(75)90109-9.
[17] Y. Benjamini and Y. Hochberg, “Controlling the false discovery rate: a practical and powerful approach to multiple testing,” Journal of the Royal Statistical Society: Series B, vol. 57, no. 1, pp. 289–300, 1995, doi: 10.1111/j.2517-6161.1995.tb02031.x.
[18] M. Kanehisa and S. Goto, “KEGG: Kyoto Encyclopedia of Genes and Genomes,” Nucleic Acids Research, vol. 28, no. 1, pp. 27–30, 2000, doi: 10.1093/nar/28.1.27.
[19] E. R. Bevan, A. M. Jones, and P. M. Hawkey, “Global epidemiology of CTX-M β-lactamases: temporal and geographical shifts in genotype,” Journal of Antimicrobial Chemotherapy, vol. 72, no. 8, pp. 2145–2155, 2017, doi: 10.1093/jac/dkx146.
[20] M. S. Ramirez and M. E. Tolmasky, “Aminoglycoside modifying enzymes,” Drug Resistance Updates, vol. 13, no. 6, pp. 151–171, 2010, doi: 10.1016/j.drup.2010.08.003.
[21] D. C. Hooper and G. A. Jacoby, “Mechanisms of drug resistance: quinolone resistance,” Annals of the New York Academy of Sciences, vol. 1354, no. 1, pp. 12–31, 2015, doi: 10.1111/nyas.12830.
[22] P. Huovinen, “Resistance to trimethoprim-sulfamethoxazole,” Clinical Infectious Diseases, vol. 32, no. 11, pp. 1608–1614, 2001, doi: 10.1086/320532.
[23] M. R. Gillings, “Integrons: past, present, and future,” Microbiology and Molecular Biology Reviews, vol. 78, no. 2, pp. 257–277, 2014, doi: 10.1128/MMBR.00056-13.
[24] Y. Li, X. Cui, X. Yang, G. Liu, and J. Zhang, “Artificial intelligence in predicting pathogenic microorganisms’ antimicrobial resistance: challenges, progress, and prospects,” Frontiers in Cellular and Infection Microbiology, vol. 14, p. 1482186, 2024, doi: 10.3389/fcimb.2024.1482186.
[25] Y. Gao, H. Li, C. Zhao, S. Li, G. Yin, and H. Wang, “Machine learning and feature extraction for rapid antimicrobial resistance prediction of Acinetobacter baumannii from whole-genome sequencing data,” Frontiers in Microbiology, vol. 14, p. 1320312, 2024, doi: 10.3389/fmicb.2023.1320312.
[26] S. G. Earle et al., “Identifying lineage effects when controlling for population structure improves power in bacterial association studies,” Nature Microbiology, vol. 1, p. 16041, 2016, doi: 10.1038/nmicrobiol.2016.41.

Details

Primary Language

English

Subjects

Image Processing, Deep Learning, Bioengineering (Other)

Journal Section

Research Article

Authors

Mesut Ersin Sönmez ^*
0000-0002-0966-9216
Türkiye

Publication Date

September 30, 2026

Submission Date

June 26, 2026

Acceptance Date

September 22, 2026

Published in Issue

Year 2026 Number: 066

IZ

https://izlik.org/JA88JL58RN

Cite

RIS / Bibtex

APA

Sönmez, M. E. (2026). Explainable artificial intelligence for genomic prediction of antimicrobial resistance in Escherichia coli: a leakage-controlled, cross-model, and biologically validated pilot framework. Journal of Scientific Reports-A, 066, 88-104. https://izlik.org/JA88JL58RN

AMA

1.Sönmez ME. Explainable artificial intelligence for genomic prediction of antimicrobial resistance in Escherichia coli: a leakage-controlled, cross-model, and biologically validated pilot framework. JSR-A. 2026;(066):88-104. https://izlik.org/JA88JL58RN

Chicago

Sönmez, Mesut Ersin. 2026. “Explainable Artificial Intelligence for Genomic Prediction of Antimicrobial Resistance in Escherichia Coli: A Leakage-Controlled, Cross-Model, and Biologically Validated Pilot Framework”. Journal of Scientific Reports-A, nos. 066: 88-104. https://izlik.org/JA88JL58RN.

EndNote

Sönmez ME (September 1, 2026) Explainable artificial intelligence for genomic prediction of antimicrobial resistance in Escherichia coli: a leakage-controlled, cross-model, and biologically validated pilot framework. Journal of Scientific Reports-A 066 88–104.

IEEE

[1]M. E. Sönmez, “Explainable artificial intelligence for genomic prediction of antimicrobial resistance in Escherichia coli: a leakage-controlled, cross-model, and biologically validated pilot framework”, JSR-A, no. 066, pp. 88–104, Sept. 2026, [Online]. Available: https://izlik.org/JA88JL58RN

ISNAD

Sönmez, Mesut Ersin. “Explainable Artificial Intelligence for Genomic Prediction of Antimicrobial Resistance in Escherichia Coli: A Leakage-Controlled, Cross-Model, and Biologically Validated Pilot Framework”. Journal of Scientific Reports-A. 066 (September 1, 2026): 88-104. https://izlik.org/JA88JL58RN.

JAMA

1.Sönmez ME. Explainable artificial intelligence for genomic prediction of antimicrobial resistance in Escherichia coli: a leakage-controlled, cross-model, and biologically validated pilot framework. JSR-A. 2026;:88–104.

MLA

Sönmez, Mesut Ersin. “Explainable Artificial Intelligence for Genomic Prediction of Antimicrobial Resistance in Escherichia Coli: A Leakage-Controlled, Cross-Model, and Biologically Validated Pilot Framework”. Journal of Scientific Reports-A, no. 066, Sept. 2026, pp. 88-104, https://izlik.org/JA88JL58RN.

Vancouver

1.Mesut Ersin Sönmez. Explainable artificial intelligence for genomic prediction of antimicrobial resistance in Escherichia coli: a leakage-controlled, cross-model, and biologically validated pilot framework. JSR-A [Internet]. 2026 Sep. 1;(066):88-104. Available from: https://izlik.org/JA88JL58RN