A Hybrid Metaheuristic based Feature Selection Framework for In-silico Mutagenicity Prediction

Öz

Mutagenicity is both a toxic risk to humans and an indicator of carcinogenicity. Hence, estimating mutagenicity in the early stages of drug design is crucial to minimize last-stage failures and withdrawals in drug discovery. Recently, in-silico methods have started to play critical and essential roles in the drug development process because they are low cost and low effort procedures. This study aims to predict mutagenicity of chemicals using in-silico methods. To achieve this goal, a two-phased flexible framework was proposed: 1) searching the effective and representative descriptors subset with Butterfly Optimization Algorithm (BOA) and Particle Swarm Optimization and 2) predicting mutagenicity of chemicals by the selected descriptor using gradient boosted tree-based ensemble methods. The study used two datasets: one including 8167 compounds for descriptor selection and modelling, and another containing 716 external compounds to validate the efficacy of our models. The datasets comprise 162 descriptors calculated using PaDEL. The results of both the cross-validation and the external data showed that descriptors reduced by nearly one-third by BOA (51 descriptors) yielded similar or slightly better predictive results than results obtained with the entire data set. The accuracy range attained by the proposed approach using BOA is approximately 91.9% to 97.91% for the external set and 83.35% to 86.47% for the test set. This research contributes that using optimization techniques for improving early drug design and minimizing risks in drug discovery can be considered as a valuable insights and advances in the field of drug toxicity prediction, based on the findings.

Anahtar Kelimeler

• Machine Learning
• Feature Selection
• Metaheuristics
• Gradient Boosting Algorithms
• Mutagenicity Prediction
• In-Silico Modelling

In-silico Mutajenite Tahmini için Hibrit Metasezgisel Tabanlı Özellik Seçimi Çerçevesi

Öz

Mutajenite hem insanlar için toksik bir risk hem de kanserojenitenin bir göstergesidir. Bu nedenle, ilaç tasarımının erken aşamalarında mutajenitenin tahmin edilmesi, ilaç keşfinde son aşama başarısızlıklarını ve geri çekilmeleri en aza indirmek için çok önemlidir. Son zamanlarda, in-silico yöntemler, düşük maliyetli ve az çaba gerektiren prosedürler olmaları nedeniyle ilaç geliştirme sürecinde kritik ve önemli roller oynamaya başlamıştır. Bu çalışma, in-silico yöntemler kullanarak kimyasalların mutajenitesini tahmin etmeyi amaçlamaktadır. Bu amaca ulaşmak için iki aşamalı esnek bir çerçeve önerilmiştir: 1) Kelebek Optimizasyon Algoritması (BOA) ve Parçacık Sürü Optimizasyonu ile etkili ve temsili değişken alt kümesinin aranması ve 2) gradyan destekli ağaç tabanlı topluluk yöntemleri kullanılarak seçilen değişkenlere göre kimyasalların mutajenitesinin tahmin edilmesi. Çalışmada iki veri kümesi kullanılmıştır: biri değişken seçimi ve modelleme için 8167 bileşik, diğeri ise modellerimizin etkinliğini doğrulamak için 716 harici bileşik içermektedir. Veri kümeleri PaDEL kullanılarak hesaplanan 162 değişkeni içermektedir. Hem çapraz doğrulama hem de harici verilerin sonuçları, BOA ile neredeyse üçte bir oranında azaltılan değişkenlerin (51 adet), tüm veri setiyle elde edilen sonuçlara benzer veya biraz daha iyi tahmin sonuçları verdiğini göstermiştir. BOA kullanılarak önerilen yaklaşımla elde edilen doğruluk aralığı harici set için yaklaşık %91,9 ila %97,91 ve test seti için %83,35 ila %86,47'dir. Bu araştırma, bulgulara dayanarak, erken ilaç tasarımını iyileştirmek ve ilaç keşfindeki riskleri en aza indirmek için optimizasyon tekniklerinin kullanılmasının, ilaç toksisitesi tahmini alanında değerli bir içgörü ve ilerleme olarak kabul edilebileceğine katkıda bulunmaktadır.

Anahtar Kelimeler

• Makine Öğrenmesi
• Özellik Seçimi
• Metasezgisel
• Gradyan Boosting Algoritmaları
• Mutajenite Tahmini
• In-Silico Modelleme

Kaynakça

1. Algamal, ZY, Qasim, MK, Lee, MH and Ali, HTM. 2020. High-dimensional QSAR/QSPR classification modelling based on improving pigeon optimization algorithm. Chemom. Intell. Lab. Syst, 206:104170, doi:10.1016/ j.chemolab.2020.104170.
2. Arora, S, Singh, S. 2019. Butterfly optimization algorithm: a novel approach for global optimization. Soft Comput. 23, 715–34 doi:10.1007/ s00500-018-3102-4.
3. Bakhtyari, N, Raitano, G, Benfenati, E, Martin, T and Young, D. 2013. Comparison of in silico models for prediction of mutagenicity. Carcinog. Ecotoxicol. Rev, 31(1):45–66, doi:10.1080/10590501.2013.763576
4. Breiman, L. 2001. Random forests. Mach. Learn, 45:5–32. doi:10.1023/ A:1010933404324.
5. Çakmak Pehlivanlı, A. and Çakmak, G. 2022. Genotoksik etkiyi belirlemeye yönelik in-silico yaklaşımlar. In Genetik Toksikoloji (Genetic Toxicology), ed. F. Ünal and D. Yüzbaşıoğlu, 475–92. Ankara: Nobel.
6. Cariello, NF, Wilson, JD, Britt, BH, Wedd, DJ, Burlinson, B and Gombar, V. 2002. Comparison of the computer programs DEREK and TOPKAT to predict bacterial mutagenicity. Mutagenesis 17(4):321-9, doi:10.1093/mutage/17.4.321.
7. Chen, T and Guestrin, C. 2016. XGBoost. In Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, 785–94, doi:10.1145/2939672.2939785.
8. Chu, CSM, Simpson, JD, O’Neill, PM and Berry, NG. 2021. Machine learning predicting Ames mutagenicity of small molecules. Journal of Molecular Graphics and Modelling, 109. doi:10.1016/j.jmgm.2021.108011.

9. Cover, T. and Hart, P. 1967. Nearest neighbor pattern classification. IEEE Trans. Inf.Theory, 13(1):21–27. doi:10.1109/TIT.1967.1053964.
10. Fan, D., Yang, H., Li, F, Sun, L, Di, P, Li, W, Tang, Y and Liu, G. 2018. In silico prediction of chemical genotoxicity using machine learning methods and structural alerts. Toxicol, 7(2): 211–20. doi:10. 1039/c7tx00259a.
11. Geurts, P, Ernst, D and Wehenkel, L. 2006. Extremely randomized trees. Mach Learn, 63:3–42. doi:10.1007/s10994-006-6226-1.
12. Greene, N, Judson, P, Langowski, J and Marchant, C. 1999. Knowledge-based expert systems for toxicity and metabolism prediction: DEREK, StAR and METEOR. SAR QSAR Environ. Res. 10 2-3, 299-314. doi:10.1080/10629369908039182.
13. Guan, D, Fan, K, Spence, I and Matthews, S. 2018. QSAR ligand dataset for modelling mutagenicity, genotoxicity, and rodent carcinogenicity. Data Br, 17:876–84, doi: 10.1016/j.dib.2018.01.077.
14. Gupta, V and Rana, P. 2019. Toxicity prediction of small drug molecules of aryl hydrocarbon receptor using a proposed ensemble model. Turkish J. Electr. Eng. Comput. Sci, 27(4): 2833–49. doi:10.3906/elk-1809-9.
15. Hansch, C. 1980. Use of quantitative structure-activity relationships (QSAR) in drug design (review). Pharmaceutical Chemistry Journal, 14. doi: 10.1007/BF00765654.
16. Hansen, K, Mika, S, Schroeter, T, Sutter, A, Laak, AT, Steger-Hartmann, T, Heinrich, N and Müller, KR. 2009. Benchmark data set for in silico prediction of Ames mutagenicity. J. Chem. Inf. Model. 49, 9, 2077–81. doi:10.1021/ci900161g.
17. Haykin, S. 2011. Neural Networks and Learning Machines. Pearson Education, 3rd ed.
18. Ho, T. 1998. The random subspace method for constructing decision forests. IEEE Trans. Pattern Anal. Mach. Intell, 20(8):832–44. doi: 10.1109/34.709601.
19. Honma M, Kitazawa, A, Cayley, A, Williams, RV, Barber, C, Hanser, T, Saiakhov, R, Chakravarti, S, Myatt, GJ, Cross, KP et. al.2019. Improvement of quantitative structure-activity relationship (QSAR) tools for predicting ames mutagenicity: Outcomes of the Ames/QSAR international challenge project. Mutagenesis, 34:41–48. doi:10. 1093/mutage/gey031.
20. Houssein, E, Honey, M, Oliva, D, Mohamed, W and Hassaballah, M. 2020. A novel hybrid harris hawks optimization and support vector machines for drug design and discovery. Comput. Chem. Eng, 133:106656. doi:10. 1016/j.compchemeng.2019.106656.
21. Ji, X, Tong, W, Liu, Z and Shi, T. 2019. Five-feature model for developing the classifier for synergistic vs. antagonistic drug combinations built by XGBoost. Front. Genet, 10(JUL):1–13. doi:10.3389/fgene.2019.00600.
22. John, L, Mahanta, HJ, Soujanya, Y, Narahari Sastry, G. 2023. Assessing machine learning approaches for predicting failures of investigational drug candidates during clinical trials. Computers in Biology and Medicine, Vol.153, 106494. doi: 10.1016/j.compbiomed.2022.106494.
23. Kazius, J, McGuire, R and Bursi, R. 2005. Derivation and validation of toxicophores for mutagenicity prediction. J. Med. Chem, 48. doi:10.1021/ jm040835a.
24. Ke, G, Meng, Q, Finley, T, Wang, T, Chen, W, Ma, W, Ye, Q and Liu, T. 2017. LightGBM: A highly efficient gradient boosting decision tree. In Proceedings of the 31st International Conference on Neural Information Processing Systems, Vol. 30, 3149–57. Curran Associates Inc.
25. Kennedy, J and Eberhart, R. 1995. Particle swarm optimization. pages 1942–48. doi:10.1109/ICNN.1995.488968.
26. Liao, Q, Yao, J and Yuan, S. 2007. Prediction of mutagenic toxicity by combination of recursive partitioning and support vector machines. Mol. Divers 11, 59–72. doi:10.1007/s11030-007-9057-5.
27. Mazzatorta, P, Tran, L, Schilter, B and Grigorov, M. 2007. Integration of structure activity relationship and artificial intelligence systems to improve in silico prediction of ames test mutagenicity. J. Chem. Inf. Model. 47, 1, 34–38. doi: 10.1021/ci600411v.
28. Mirjalili, S and Lewis, A. 2013. S-shaped versus V-shaped transfer functions for binary particle swarm optimization. Swarm Evol. Comput, 9:1–14. 10.1016/j.swevo.2012.09.002.
29. Mitchell, T. 1997. Machine Learning. McGraw-Hill. New York
30. Moorthy, N, Kumar, S and Poongavanam, V. 2017. Classification of carcinogenic and mutagenic properties using machine learning method. Comput. Toxicol, 3:33–43. doi: 10.1016/j.comtox.2017.07.002.
31. Raghavan, N, Amaratunga, D, Nie, AY and McMillian, M. 2005. Class prediction in toxicogenomics, Journal of Biopharmaceutical Statistics, 15:2, 327-41, doi: 10.1081/BIP-200048836
32. Rifaioglu, AS, Atas, H, Martin, MJ, Cetin-Atalay, R, Atalay,V and Doǧan, T. 2019. Recent applications of deep learning and machine intelligence on in silico drug discovery: Methods, tools and databases. Brief. Bioinform, 20(5):1878–1912. doi: 10.1093/bib/bby061.
33. Seal, A, Passi, A, Jaleel, U, Wild, D and Consortium, O. 2012. In-silico predictive mutagenicity model generation using supervised learning approaches. J. Cheminform. 4(1):10. doi:10.1186/1758-2946-4-10.
34. Sharma, A, Kumar, R, Varadwaj, P, Ahmad, A and Ashraf, G. 2011. A comparative study of support vector machine, artificial neural network and bayesian classifier for mutagenicity prediction. Interdiscip. Sci. Comput. Life Sci, 3(3):232–239. doi:10.1007/s12539-011-0102-9.
35. Stevens, S.S. 1986. Psychophysics: Introduction to Its Perceptual, Neural and Social Prospects. 1st ed. Routledge. doi.org/10.4324/9781315127675
36. Subaş, N and Çakmak Pehli̇vanlı, A. 2020. İkili parçacık sürü optimizasyonu ve destek vektör makinelerinin hibrit kullanımı ile ilaç keşfi için özellik seçimi. Gümüşhane Üniv. Fen Bilim. Enst. Derg., 11:169–78. doi:10. 17714/gumusfenbil.776329.
37. Tran, T T V, Surya Wibowo, A, Tayara, H and Chong, KT. 2023. Artificial intelligence in drug toxicity prediction: Recent advances, challenges, and future perspectives. Journal of Chemical Information and Modeling, 63(9):2628–43. doi: 10.1021/acs.jcim.3c00200.
38. Toropov, AA, Toropova, AP, Raska, I, Leszczynska, D, Leszczynski, J. 2014. Comprehension of drug toxicity: Software and databases. Computers in Biology and Medicine, 45: 20-25. doi: 10.1016/j.compbiomed.2013.11.013.
39. Tubishat, M, Alswaitti, M, Mirjalili, S, Al-Garage, M, Alrashdan, M and Rana, T. 2020. Dynamic butterfly optimization algorithm for feature selection. IEEE Access, 8:194303–14. doi:10.1109/access.2020.3033757.
40. Vapnik, V. 1995. The Nature of Statistical Learning Theory. Springer-Verlag, Berlin, Heidelberg.
41. Webb, SJ, Hanser, T, Howlin, B, Krause, P and Vessey, J. 2014a. Feature combination networks for the interpretation of statistical machine learning models: Application to Ames mutagenicity. J. Cheminform, 6(1):8. doi: 10.1186/1758-2946-6-8.
42. Webb, SJ, Hanser, T, Howlin, B, Krause, P and Vessey, J. 2014b. Interpretable Ames mutagenicity predictions using statistical learning techniques. In Handbook of abstracts, 6th Joint Sheffield Conference on Chemoinformatics. Qsar2012, 3–4.
43. White, A, Mueller, R, Gallavan, R, Aaron, A and Wilson, A. 2003. A multiple in silico program approach for the prediction of mutagenicity from chemical structure. Mutat. Res. - Genet. Toxicol. Environ. Mutagen, 539:77–89. doi:10.1016/S1383-5718(03)00135-9.
44. Wichard, J.D. 2017. In silico prediction of genotoxicity. Food and Chemical Toxicology, 106(Pt B):595-599. doi: 10.1016/j.fct.2016.12.013.
45. Xu, C, Cheng, F, Chen, L, Du, Z, Li, W, Liu, G, Lee, PW and Tang,Y. 2012. In silico prediction of chemical names mutagenicity. Journal of Chemical Information and Modeling, 52(11):2840–47. doi:10.1021/ci300400a.
46. Yang, Z, Thomson, R, Mcneil, P and Esnouf, R. 2005. Ronn: the bio-basis function neural network technique applied to the detection of natively disordered regions in proteins. Bioinforma. Orig. Pap, 21(16):3369–3376. doi:10.1093/bioinformatics/bti534.
47. Yap, C. 2010. Padel-descriptor: an open-source software to calculate molecular descriptors and fingerprints. J. Comput. Chem, 32(4):1466–74. doi:10.1002/jcc.21707.
48. Zhang, H, Yu, P, Zhang, TG, Kang, YL, Zhao, X, Li, YY, He, JH and Zhang, J. 2015. In silico prediction of drug-induced myelotoxicity by using naïve Bayes method. Mol. Divers, 19(4): 945-53. doi: 10.1007/s11030-015-9613-3.
49. Zhang, H, Yu, P, Xiang, ML, Li, XB, Kong, WB, Ma, JY, Wang, JL, Zhang, JP and Zhang, J. 2016. Prediction of drug-induced eosinophilia adverse effect by using SVM and naïve Bayesian approaches. Med. Biol. Eng. Comput, 54(2–3):361–9. doi: 10.1007/s11517-015-1321-8.
50. Zhang, H, Kang, YL, Zhu, YY, Zhao, KX, Liang, JY, Ding, L, Zhang,TG and Zhang, J. 2017. Novel naïve Bayes classification models for predicting the chemical Ames mutagenicity. Toxicol. Vitr, 41:56–63. doi: 10.1016/j.tiv.2017.02.016.
51. Zhang, J, Mucs, D, Norinder, U and Svensson, F. 2019. LightGBM: An effective and scalable algorithm for prediction of chemical toxicity-application to the Tox21 and mutagenicity data sets. J. Chem. Inf. Model., 59(10):4150–58. doi: 10.1021/acs.jcim.9b00633.
52. Zheng, M, Liu, Z, Xue, C, Zhu, W, Chen, K, Luo, X and Jiang, H. 2006. Mutagenic probability estimation of chemical compounds by a novel molecular electrophilicity vector and support vector machine. Bioinformatics. 22(17):2099-106. doi: 10.1093/bioinformatics/btl352.