İLAÇ LD50 DEĞERİNİ TAHMİN ETMEK İÇİN ÇOK DEĞİŞKENLİ BİR İNTERPOLASYON YAKLAŞIMI

Öz

Amaç: Bu çalışmanın amacı, beş farklı toksikolojik sonucu dikkate alarak farelerde ilaçların oral median letal doz (LD50) değerlerini doğru bir şekilde tahmin edebilen, niceliksel yapı-toksisite ilişkisine (QSTR) dayalı çok değişkenli bir interpolasyon yöntemi geliştirmektir. Gereç ve Yöntem: Farklı farmakolojik sınıflara ait 319 ilaca ait LD50 değerlerini içeren kapsamlı bir veri seti kullanılarak matematiksel bir model oluşturuldu. Farmasötiklerin LD50 değerlerinin aralığını tahmin edebilen bir polinom model geliştirdik. İki değişkenli polinom interpolasyon adı verilen bir teknik kullanarak bunu gerçekleştirdik. Bu yöntem, bir polinom denklemi kullanarak iki boyutlu bir uzayda herhangi bir noktadaki bir fonksiyonun değerlerini tahmin etmemizi sağladı. Sonuç ve Tartışma: Elde edilen model, yeni veya denenmemiş ilaçlar için LD50 değerlerini tahmin etme yeteneğini gösterdi ve bu nedenle ilaç geliştirme sürecinin erken aşamalarında değerli bir araç olarak kullanılabilir. Değerlendirmemize göre, model genel başarı oranı olarak %86,73 olarak bulundu. LD50 değerinin belirlenmesinde kullanılan geleneksel deneysel yöntemlere kıyasla, bu yenilikçi yaklaşım zaman ve maliyet açısından avantajlı olup hayvan deneylerinin gerekliliğini azaltmaktadır. Modelimiz ilaç güvenliğini artırabilir, doz rejimlerini optimize edebilir ve ön klinik çalışmalar ve ilaç geliştirme sürecinde karar verme süreçlerine yardımcı olabilir. Bu yaklaşım, ön akut toksisite değerlendirmeleri için güvenilir ve etkili bir yöntem sunmuştur.

Anahtar Kelimeler

• Çok değişkenli interpolasyon
• LD50
• matematiksel toksikoloji
• polinom interpolasyonu
• veri analizi

A MULTIVARIATE INTERPOLATION APPROACH FOR PREDICTING DRUG LD50 VALUE

Öz

Objective: The present study aimed to develop a multivariate interpolation based on the quantitative structure-toxicity relationship (QSTR) that can accurately predict the oral median lethal dose (LD50) values of drugs in mice by considering five different toxicologic endpoints. Material and Method: A mathematical model was created using a comprehensive dataset comprising LD50 values from 319 pharmaceuticals belonging to various pharmacological classes. We developed a polynomial model that can predict the range of LD50 values for pharmaceuticals. We employed a technique called two-variable polynomial interpolation. This method allowed us to estimate the approximate values of a function at any point within a two-dimensional (2D) space by utilizing a polynomial equation. Result and Discussion: The resulting model demonstrated the ability to predict LD50 values for new or untested drugs, rendering it a valuable tool in the early stages of drug development. The Ghose-Crippen-Viswanadhan octanol-water partition coefficient (ALogP) and Molecular Weight (MW) were selected as suitable descriptors for building the best QSAR model. Based on our evaluation, the model achieved an overall success rate of 86.73%. Compared to traditional experimental methods for LD50 determination, this innovative approach offers time and cost efficiency while reducing animal testing requirements. Our model can improve drug safety, optimize dosage regimens, and assist decision-making processes during preclinical studies and drug development. This approach provided a reliable and efficient method for preliminary acute toxicity assessments.

Anahtar Kelimeler

• Data analysis
• LD50
• mathematical toxicology
• multivariate interpolation
• polynomial interpolation

Kaynakça

1. 1. United Nations Web site. (2007). Globally Harmonized System of Classification and Labelling of Chemicals (GHS), ST/SG/AC.10/30/Rev.2. https://unece.org/. Access date: 03.05.2023.
2. 2. Food and Drug Administration (FDA).
3. 3. Akkaya, H., Kelleci Çelik, F. (2021). Hayvan Deneylerine Etik Açıdan Bakış. Atatürk Üniversitesi Yayınları, Erzurum, p. 75.
4. 4. Gadaleta, D., Vuković, K., Toma, C., Lavado, G.J., Karmaus, A.L., Mansouri, K., Kleinstreuer, N.C., Benfenati, E., Roncaglioni, A. (2019). SAR and QSAR modeling of a large collection of LD50 rat acute oral toxicity data. Journal of Cheminformatics, 11(1), 58. [CrossRef]
5. 5. Karaduman, G., Kelleci Çelik, F. (2023). 2D-Quantitative structure-activity relationship modeling for risk assessment of pharmacotherapy applied during pregnancy. Journal of Applied Toxicology, 43(10), 1436-1446. [CrossRef]
6. 6. Rasulev, B., Kusić, H., Leszczynska, D., Leszczynski, J., Koprivanac, N. (2010). QSAR modeling of acute toxicity on mammals caused by aromatic compounds: The case study using oral LD50 for rats. Journal of Environmental Monitoring, 12(5), 1037-1044. [CrossRef]
7. 7. Ruiz, P., Begluitti, G., Tincher, T., Wheeler, J., Mumtaz, M. (2012). Prediction of acute mammalian toxicity using QSAR methods: a case study of sulfur mustard and its breakdown products. Molecules, 17(8), 8982-9001. [CrossRef]
8. 8. Lapenna, S., Gatnik, M.F., Worth, A.P. (2010). Review of QSAR models and software tools for predicting acute and chronic systemic toxicity. Publications Office of the European Union, Luxembourg, JRC61930, 1-35.

9. 9. Schaper, M.M., Thompson, R.D., Weil, C.S. (1994). Computer programs for calculation of median effective dose (LD50 or ED50) using the method of moving average interpolation. Archives Toxicology, 68, 332-337 [CrossRef]
10. 10. Lin, Z., Chou, W.C. (2022). Machine learning and artificial intelligence in toxicological sciences. Toxicological Sciences, 189(1), 7-19. [CrossRef]
11. 11. Lane, T.R., Harris, J., Urbina, F., Ekins, S. (2023). Comparing LD50/LC50 machine learning models for multiple species. ACS Chemical Health and Safety, 30(2), 83-97. [CrossRef]
12. 12. Burden, R.L., Faires, J.D. (1997). Numerical analysis (6th ed.). Brooks/Cole Pub.
13. 13. The European Chemicals Agency (ECHA). Web site. From https://echa.europa.eu/. Access date: 03.05.2023.
14. 14. European Food Safety Authority (EFSA). Web site. From https://www.efsa.europa.eu/en. Access date: 03.05.2023.
15. 15. National Library of Medicine (NLM). Web site. From https://www.nlm.nih.gov/. Access date: 03.05.2023.
16. 16. Kim, S., Chen, J., Cheng, T., Gindulyte, A., He, J., He, S., Li, Q., Shoemaker, B.A., Thiessen, P.A., Yu, B., Zaslavsky, L., Zhang, J., Bolton, E.E. (2023). PubChem 2023 update. Nucleic Acids Research, 51(D1), D1373-D1380. [CrossRef]
17. 17. The United States Environmental Protection Agency (U.S. EPA) Web site. (2020). User's guide for T. E. S. T. (Toxicity Estimation Software Tool) version 5.1 a java application to estimate toxicities and physical properties from molecular structure. From https://www.epa.gov/chemical-research/toxicity-estimation-software-tool-test. Accessed date: 03.05.2023.
18. 18. Miličević, A., Šinko, G. (2022). Evaluation of the key structural features of various butyrylcholinesterase inhibitors using simple molecular descriptors. Molecules, 27(20), 6894. [CrossRef]
19. 19. OECD (2010), Test No. 417: Toxicokinetics, OECD guidelines for the testing of chemicals, Section 4, OECD Publishing, Paris. [CrossRef]
20. 20. Akturk, S.O., Tugcu, G., Sipahi, H. (2022). Development of a QSAR model to predict comedogenic potential of some cosmetic ingredients. Computational Toxicology, 21, 100207. [CrossRef]
21. 21. Bojanov, B., Xu, Y. (2003). On polynomial interpolation of two variables. Journal of Approximation Theory, 120(2), 267-282. [CrossRef]
22. 22. Hust, J.G., McCarty, R.D. (1967). Curve-fitting techniques and applications to thermodynamics, Cryogenics, 7(1), 200-206. [CrossRef]
23. 23. Mehari, Y. (2017). Easy way to find multivariate interpolation. International Journal of Emerging Trends in Science and Technology, 4(5), 5189-5193.
24. 24. Karaduman, G., Yang, M. (2022). An alternative method for SPP with full rank (2,1)-block matrix and nonzero right-hand side vector. Turkish Journal of Mathematics, 46(4), 1330-1341. [CrossRef]
25. 25. OECD (2017). Guidance Document on the Validation of (Quantitative) Structure-Activity Relationship [(Q)Sar] Models. In: OECD Series on Testing and Assessment. OECD Publishing, Paris, 1-154. [CrossRef]
26. 26. Demisse, G.B., Tadesse, T., Bayissa, Y. (2017). Data mining attribute selection approach for drought modeling: A case study for Greater Horn of Africa. International Journal of Data Mining and Knowledge Management Process, 7(4), 1-16. [CrossRef]
27. 27. Kelleci Çelik, F., Karaduman, G. (2022). In silico QSAR modeling to predict the safe use of antibiotics during pregnancy. Drug and Chemical Toxicology, 46(3), 1-10. [CrossRef]
28. 28. Devillers, J. (2004). Prediction of mammalian toxicity of organophosphorus pesticides from QSTR modeling. SAR and QSAR in Environmental Research, 15(5-6), 501-510. [CrossRef]
29. 29. Abraham, M.H., Grellier, P.L., Kamlet, M.J., Doherty, R.M., Taft, R.W., Abboud, J.L.M. (1989). The use of scales of hydrogen-bond acidity and basicity in organic chemistry. Revista Portuguesa de Química, 31, 85.
30. 30. OECD (2022), Test No. 425: Acute Oral Toxicity: Up-and-Down Procedure, OECD Guidelines for the Testing of Chemicals, Section 4, OECD Publishing, Paris. [CrossRef]
31. 31. Zhu, H., Martin, T.M., Ye, L., Sedykh, A., Young, D.M., Tropsha, A. (2009). Quantitative structure activity relationship modeling of rat acute toxicity by oral exposure. Chemical Research in Toxicology, 22(12), 1913-1921. [CrossRef]