Detection of new candidate compounds against four antibiotic targets using explainable artificial intelligence by molecular fingerprints

Year 2023, Volume: 7 Issue: 2, 47 - 52, 19.12.2023

Kevser Kübra Kırboğa Naeem Abdul Ghafoor Ömür Baysal

Abstract

Antibiotic resistance is a threat that renders bacteria ineffective against antibiotics and makes it difficult to treat infections. Therefore, finding new target compounds is essential in discovering and developing new antibiotics. In this study, we developed an artificial intelligence algorithm that can predict and explain the pIC50 values for four antibiotic targets (Penicillin Binding Proteins (PB), β-Lactamase (BL), DNA Gyrase (DG), and Dihydrofolate Reductase(DR)). The algorithm uses molecular fingerprints of the molecules to predict the pIC50 values using the random forest regression method. We created the algorithm in a transparent and interpretable way. We used permutation feature importance (PFI) and Shapley explanations methods to identify the different molecular fingerprints that have the most influence on the pIC50 values. The results obtained from these methods show that different molecular fingerprints are essential for different antibiotic targets. According to the permutation importance results, KRFPC1646 (number of hydrogen bond donors of the compound) for BL and DR targets; 579 (a substructure with 5 bonded radius around the atom) for DG target; SubFPC182 (number of aromatic rings in the molecule) for PB target, are important fingerprints. With explainable artificial intelligence (XAI) (SHAP), KRFPC1646 (the number of hydrogen bond donors of the compound) for BL; KRFPC4274 (C1CCCCC1) for DR; 401 (C1CCCCC1) for DG; SubFPC182 (number of aromatic rings in the molecule) were determined as important fingerprints for PB. These results demonstrate the effectiveness and potential of using molecular fingerprints with explainable artificial intelligence to find new antibiotic candidates.

Keywords

antibiotics, explainable artificial intelligence, shapley explanations, machine learning

Supporting Institution

TUBITAK

Project Number

122E082

Thanks

This study emerged from the TUBITAK 1002, “Developing a Machine Learning-Based Bioinformatics Framework for the Identification of New Antibacterial Agents, 122E082”.

References

• [1] J. M. Stokes et al., "A Deep Learning Approach to Antibiotic Discovery," Cell, vol. 180, no. 4, pp. 688-702.e13, 2020/02/20/ 2020, doi: https://doi.org/10.1016/j.cell.2020.01.021.
• [2] E. D. Brown and G. D. Wright, "Antibacterial drug discovery in the resistance era," Nature, vol. 529, no. 7586, pp. 336-343, 2016.
• [3] P. E. W. Trusts. "Five-year analysis shows continued deficiencies in antibiotic development." https://www.pewtrusts.org/en/research-and-analysis/data-visualizations/2019/five-year-analysis-shows-continued-deficiencies-in-antibiotic-development (accessed.
• [4] J. O'Neill. "Antimicrobial Resistance:Tackling a crisis for the health and wealth of nations." https://www.ecdc.europa.eu/en/publications-data/ecdcemea-joint-technical-report-bacterial-challenge-time-react (accessed 29.05, 2023).
• [5] A. P. Ball et al., "Future trends in antimicrobial chemotherapy: expert opinion on the 43rd ICAAC," (in eng), J Chemother, vol. 16, no. 5, pp. 419-36, Oct 2004, doi: 10.1179/joc.2004.16.5.419.
• [6] R. P. Bax et al., "Antibiotic resistance - what can we do?," Nature Medicine, vol. 4, no. 5, pp. 545-546, 1998/05/01 1998, doi: 10.1038/nm0598-545.
• [7] A. R. Coates and Y. Hu, "Novel approaches to developing new antibiotics for bacterial infections," (in eng), Br J Pharmacol, vol. 152, no. 8, pp. 1147-54, Dec 2007, doi: 10.1038/sj.bjp.0707432.
• [8] K. Chaibi et al., "What to Do with the New Antibiotics?," ANTIBIOTICS-BASEL, vol. 12, no. 4, APR 2023, Art no. 654, doi: 10.3390/antibiotics12040654.
• [9] O. V. Kisil, N. I. Gabrielyan, and V. V. Maleev, "Antibiotic resistance - what can be done? A review," TERAPEVTICHESKII ARKHIV, vol. 95, no. 1, pp. 90-95, 2023, doi: 10.26442/00403660.2023.01.202040.
• [10] K. K. Kırboğa, S. Abbasi, and E. U. Küçüksille, "Explainability and white box in drug discovery," Chemical Biology & Drug Design, vol. n/a, no. n/a, doi: https://doi.org/10.1111/cbdd.14262.
• [11] L. David et al., "Artificial Intelligence and Antibiotic Discovery," Antibiotics, vol. 10, no. 11, p. 1376, 2021. [Online]. Available: https://www.mdpi.com/2079-6382/10/11/1376.
• [12] K. K. Kırboğa and E. U. Küçüksille, "Perspectives on Computer Aided Drug Discovery," (in en scheme="ISO639-1"), 11, Review Articles 2023, doi: https://dergipark.org.tr/en/pub/dufed/issue/70232/1103457.
• [13] B. Cunha, L. Fonseca, and C. Calado, "Simultaneous elucidation of antibiotics mechanism-of-action and potency with high-throughput fourier-transform Infrared spectroscopy and machine-learning," App. Microb. Biot, vol. 105, pp. 1269-1286, 2021.
• [14] S. Zoffmann et al., "Machine learning-powered antibiotics phenotypic drug discovery," Scientific reports, vol. 9, no. 1, p. 5013, 2019.
• [15] J. M. Stokes et al., "A deep learning approach to antibiotic discovery," Cell, vol. 180, no. 4, pp. 688-702. e13, 2020.
• [16] N. Parvaiz, F. Ahmad, W. Yu, A. D. MacKerell Jr, and S. S. Azam, "Discovery of beta-lactamase CMY-10 inhibitors for combination therapy against multi-drug resistant Enterobacteriaceae," PLoS One, vol. 16, no. 1, p. e0244967, 2021.
• [17] M.-N. Hamid and I. Friedberg, "Identifying antimicrobial peptides using word embedding with deep recurrent neural networks," Bioinformatics, vol. 35, no. 12, pp. 2009-2016, 2019.
• [18] A. Badura, J. Krysiński, A. Nowaczyk, and A. Buciński, "Application of artificial neural networks to prediction of new substances with antimicrobial activity against Escherichia coli," Journal of Applied Microbiology, vol. 130, no. 1, pp. 40-49, 2021.
• [19] N. Macesic, O. J. Bear Don't Walk IV, I. Pe'er, N. P. Tatonetti, A. Y. Peleg, and A.-C. Uhlemann, "Predicting phenotypic polymyxin resistance in Klebsiella pneumoniae through machine learning analysis of genomic data," Msystems, vol. 5, no. 3, pp. e00656-19, 2020.
• [20] E. N. Grafskaia et al., "Discovery of novel antimicrobial peptides: A transcriptomic study of the sea anemone Cnidopus japonicus," Journal of Bioinformatics and Computational Biology, vol. 16, no. 02, p. 1840006, 2018.
• [21] X. Su, J. Xu, Y. Yin, X. Quan, and H. Zhang, "Antimicrobial peptide identification using multi-scale convolutional network," BMC bioinformatics, vol. 20, no. 1, pp. 1-10, 2019.
• [22] M. Tharmakulasingam, B. Gardner, R. L. Ragione, and A. Fernando, "Explainable Deep Learning Approach for Multilabel Classification of Antimicrobial Resistance With Missing Labels," IEEE Access, vol. 10, pp. 113073-113085, 2022, doi: 10.1109/ACCESS.2022.3216896.
• [23] M. I. Oladunjoye, O. O. Obe, and O. D. Alowolodu, A deep neural network for the identification of lead molecules in antibiotics discovery (Explainable Artificial Intelligence in Medical Decision Support Systems). 2022, pp. 381-400.
• [24] B. Cánovas-Segura et al., "Exploring Antimicrobial Resistance Prediction Using Post-hoc Interpretable Methods," in Artificial Intelligence in Medicine: Knowledge Representation and Transparent and Explainable Systems, Cham, M. Marcos et al., Eds., 2019// 2019: Springer International Publishing, pp. 93-107.
• [25] A. Gaulton et al., "The ChEMBL database in 2017," Nucleic Acids Research, vol. 45, no. D1, pp. D945-D954, 2017-01-04 2017, doi: 10.1093/nar/gkw1074.
• [26] T. Liu, Y. Lin, X. Wen, R. N. Jorissen, and M. K. Gilson, "BindingDB: a web-accessible database of experimentally determined protein-ligand binding affinities," (in eng), Nucleic Acids Res, vol. 35, no. Database issue, pp. D198-201, Jan 2007, doi: 10.1093/nar/gkl999.
• [27] S. Kim et al., "PubChem Substance and Compound databases," (in eng), Nucleic Acids Res, vol. 44, no. D1, pp. D1202-13, Jan 4 2016, doi: 10.1093/nar/gkv951.
• [28] aporia-ai. "Permutation-importance." Github. https://github.com/aporia-ai/Permutation-importance/blob/main/Permutation%20importance/RegressionTask_carl_house.ipynb (accessed 23.05.2023, 2023).
• [29] A. Altmann, L. Toloşi, O. Sander, and T. Lengauer, "Permutation importance: a corrected feature importance measure," Bioinformatics, vol. 26, no. 10, pp. 1340-1347, 2010.
• [30] F. Pedregosa et al., "Scikit-learn: Machine Learning in Python," J. Mach. Learn. Res., vol. 12, no. null, pp. 2825–2830, 2011.
• [31] L. S. Shapley, "17. A Value for n-Person Games," in Contributions to the Theory of Games (AM-28), Volume II, K. Harold William and T. Albert William Eds. Princeton: Princeton University Press, 1953, pp. 307-318.
• [32] S. Lundberg and S.-I. Lee, A Unified Approach to Interpreting Model Predictions. 2017.
• [33] R. Mitchell, E. Frank, and G. Holmes, "GPUTreeShap: massively parallel exact calculation of SHAP scores for tree ensembles," (in eng), PeerJ Comput Sci, vol. 8, p. e880, 2022, doi: 10.7717/peerj-cs.880.
• [34] C. Molnar, "Interpretable Machine Learning," Self published, 2020. [Online]. Available: https://christophm.github.io/interpretable-ml-book/.
• [35] S. M. Lundberg, G. G. Erion, and S.-I. Lee, "Consistent Individualized Feature Attribution for Tree Ensembles," 2019-03-07T00:06:09 2019.
• [36] A. Capecchi, D. Probst, and J.-L. Reymond, "One molecular fingerprint to rule them all: drugs, biomolecules, and the metabolome," Journal of Cheminformatics, vol. 12, no. 1, p. 43, 2020/06/12 2020, doi: 10.1186/s13321-020-00445-4.
• [37] R. L. Apodaca. "Computing Extended Connectivity Fingerprints." https://depth-first.com/articles/2019/01/11/extended-connectivity-fingerprints/ (accessed.
• [38] C. M. Consulting. "Examples of Fingerprint and Descriptors." https://www.cambridgemedchemconsulting.com/resources/hit_identification/examples_descriptors.php (accessed.
• [39] R. C. Glem, A. Bender, C. H. Arnby, L. Carlsson, S. Boyer, and J. Smith, "Circular fingerprints: flexible molecular descriptors with applications from physical chemistry to ADME," (in eng), IDrugs, vol. 9, no. 3, pp. 199-204, Mar 2006.
• [40] N. Suvannang et al., "Probing the origin of estrogen receptor alpha inhibition via large-scale QSAR study," RSC Advances, 10.1039/C7RA10979B vol. 8, no. 21, pp. 11344-11356, 2018, doi: 10.1039/C7RA10979B.
• [41] C. Phanus-Umporn, W. Shoombuatong, V. Prachayasittikul, N. Anuwongcharoen, and C. Nantasenamat, "Privileged substructures for anti-sickling activity via cheminformatic analysis," (in eng), RSC Adv, vol. 8, no. 11, pp. 5920-5935, Feb 2 2018, doi: 10.1039/c7ra12079f.