Protein Folding Problem: A New Approach with Evolutionary Algorithms

Öz

AlphaFold is transforming the field of structural biology by predicting three-dimensional (3D) structures from protein sequences. This remarkable achievement has even led to claims that the protein folding problem has been 'solved.' However, the protein folding problem encompasses more than just structure prediction from sequences. Currently, it is unknown whether the AlphaFold revolution will help solve other issues related to protein folding. In this study, we evaluate AlphaFold's ability to predict the impact of single mutations on protein stability (ΔΔG) and function. To explore this question, we subtract the pLDDT and metrics from AlphaFold predictions before and after a single mutation in a protein, correlating the predicted changes with experimentally known ΔΔG values. Additionally, we correlate the same AlphaFold pLDDT metrics with fluorescence levels experimentally tested using a large-scale GFP mutant data set, which indicates the effect of single mutations on the structure. We find very weak or no correlation between AlphaFold output metrics and protein stability or fluorescence change. Our results suggest that AlphaFold may not be immediately applicable to other protein folding problems or applications.

Anahtar Kelimeler

• Protein Stability
• Protein Structure Modeling
• Structural Bioinformatics
• Ensemble Classifiers
• AlphaFold

Protein Katlanmasının Çözülmesi: Evrimsel Algoritmalar ve Yapay Zekanın Gücüyle Yenilikçi Yaklaşımlar

Öz

AlphaFold, protein dizisinden üç boyutlu (3D) yapı tahmini yaparak yapısal biyoloji alanını değiştirmektedir. Bu olağanüstü başarı, protein katlanma probleminin "çözüldüğü" iddialarına bile yol açmıştır. Ancak protein katlanma problemi, diziden yapı tahmininden daha fazlasını içermektedir. Şu anda, AlphaFold devriminin protein katlanma ile ilgili diğer sorunları çözmeye yardımcı olup olmayacağı bilinmemektedir. Bu çalışmada, AlphaFold'un tekli mutasyonların protein stabilitesi (ΔΔG) ve fonksiyonuna etkisini tahmin etme yeteneğini değerlendiriyoruz. Bu soruyu incelemek için, AlphaFold tahminlerinden pLDDT ve metriklerini bir proteinde tekli mutasyondan önce ve sonra çıkarıyoruz ve tahmin edilen değişikliği deneysel olarak bilinen ΔΔG değerleriyle ilişkilendiriyoruz. Ayrıca, aynı AlphaFold pLDDT metriklerini yapı üzerindeki tekli mutasyonun etkisi ile büyük ölçekli bir GFP mutant veri setini kullanarak deneysel olarak test edilen floresans seviyeleri ile ilişkilendiriyoruz. AlphaFold çıktı metrikleri ile protein stabilitesi veya floresan değişimi arasında çok zayıf veya hiçbir korelasyon bulamıyoruz. Sonuçlarımız, AlphaFold'un diğer protein katlanma problemleri veya uygulamalarına hemen uygulanamayabileceğini göstermektedir.

Anahtar Kelimeler

• AlphaFold;
• Protein Stabilitesi;
• Derin Öğrenme;
• Şablon Tabanlı Yaklaşımlar
• Protein Yapısı Modelleme;
• Markov Durum Modelleri

Kaynakça

1. [1] H. İşci, Z. Baykara, ve B. Tülüce, “Bulanık TOPSIS ve Bulanık AHP Yöntemleri ile Risk Analizi Örneği”, ALKÜ Fen Bilimleri Dergisi, c. 6, sy. 1, ss. 28–45, 2024, doi: 10.46740/alku.1316669.
2. [2] B. Kuhlman and P. Bradley, "Advances in protein structure prediction and design," Nature Reviews Molecular Cell Biology, vol. 20, no. 11, pp. 681–697, 2019, doi: 10.1038/s41580-019-0163-x.
3. [3] F. Chiti and C. M. Dobson, "Protein misfolding, functional amyloid, and human disease," Annual Review of Biochemistry, vol. 75, pp. 333–366, 2006, doi: 10.1146/annurev.biochem.75.101304.123901.
4. [4] W. Kühlbrandt, "The resolution revolution," Science, vol. 343, no. 6178, pp. 1443–1444, 2014, doi: 10.1126/science.1251652.
5. [5] Y. Shi, "A glimpse of structural biology through X-ray crystallography," Cell, vol. 159, no. 5, pp. 995–1014, 2014, doi: 10.1016/j.cell.2014.10.051.
6. [6] K. A. Dill and J. L. MacCallum, "The protein-folding problem, 50 years on," Science, vol. 338, no. 6110, pp. 1042–1046, 2012, doi: 10.1126/science.1219021.
7. [7] C. B. Anfinsen, "Principles that govern the folding of protein chains," Science, vol. 181, no. 4096, pp. 223–230, 1973, doi: 10.1126/science.181.4096.223.
8. [8] J. Jumper et al., "Highly accurate protein structure prediction with AlphaFold," Nature, vol. 596, no. 7873, pp. 583–589, 2021, doi: 10.1038/s41586-021-03819-2.

9. [9] A. W. Senior et al., "Improved protein structure prediction using potentials from deep learning," Nature, vol. 577, no. 7792, pp. 706–710, 2020, doi: 10.1038/s41586-019-1923-7.
10. [10] A. Kryshtafovych, T. Schwede, K. Topf, K. Fidelis, and J. Moult, "Critical assessment of methods of protein structure prediction (CASP)—Round XIV," Proteins, vol. 89, no. 12, pp. 1607–1617, 2021, doi: 10.1002/prot.26237.
11. [11] E. Callaway, "‘It will change everything’: DeepMind’s AI makes gigantic leap in solving protein structures," Nature, vol. 588, no. 7837, pp. 203–204, 2020, doi: 10.1038/d41586-020-03348-4.
12. [12] M. Hussain et al., "Machine learning based efficient prediction of positive cases of waterborne diseases," BMC Medical Informatics and Decision Making, vol. 23, no. 1, p. 11, 2023, doi: 10.1186/s12911-022-02095-0.
13. [13] J. Pereira et al., "High-accuracy protein structure prediction in CASP14," Proteins, vol. 89, no. 12, pp. 1687–1699, 2021, doi: 10.1002/prot.26171.
14. [14] M. Varadi et al., "AlphaFold Protein Structure Database: massively expanding the structural coverage of protein-sequence space with high-accuracy models," Nucleic Acids Research, vol. 50, no. D1, pp. D439–D444, 2022, doi: 10.1093/nar/gkab1061.
15. [15] M. A. Pak, K. M. Markey, S. Vastermark, and A. Biegert, "Using AlphaFold to predict the impact of single mutations on protein stability and function," bioRxiv, 2021, doi: 10.1101/2021.09.19.460937.
16. [16] Y. Dehouck et al., "Fast and accurate predictions of protein stability changes upon mutations using statistical potentials and neural networks: PoPMuSiC-2.0," Bioinformatics, vol. 25, no. 19, pp. 2537–2543, 2009, doi: 10.1093/bioinformatics/btp445.
17. [17] L. Montanucci, P. Fariselli, and R. Casadio, "DDGun: An untrained method for the prediction of protein stability changes upon single and multiple point variations," BMC Bioinformatics, vol. 20, no. S14, p. 335, 2019, doi: 10.1186/s12859-019-2923-1.
18. [18] P. Gainza et al., "Deciphering interaction fingerprints from protein molecular surfaces using geometric deep learning," Nature Methods, vol. 17, no. 2, pp. 184–192, 2020, doi: 10.1038/s41592-019-0666-6.
19. [19] M. Hussain et al., "Machine learning-driven approach for a COVID-19 warning system," Electronics, vol. 11, no. 23, p. 3875, 2022, doi: 10.3390/electronics11233875.
20. [20] B. Li, Y. Yang, and Z. Zhang, "Characterization of the impact of missense mutations on proteins through machine learning," BMC Genomics, vol. 22, no. S3, p. 297, 2021, doi: 10.1186/s12864-021-07567-8.
21. [21] V. E. Gray, S. R. Hause, and D. M. Fowler, "Elucidating the molecular determinants of Aβ aggregation with deep mutational scanning," G3: Genes, Genomes, Genetics, vol. 9, no. 10, pp. 3683–3689, 2019, doi: 10.1534/g3.119.400473.
22. [22] M. A. Cifci, S. Hussain, and P. J. Canatalay, "Hybrid deep learning approach for accurate tumor detection in medical imaging data," Diagnostics, vol. 13, no. 6, p. 1025, 2023, doi: 10.3390/diagnostics13061025.
23. [23] D. M. Fowler and S. Fields, "Deep mutational scanning: A new style of protein science," Nature Methods, vol. 11, no. 8, pp. 801–807, 2014, doi: 10.1038/nmeth.3027.
24. [24] B. Steinberg, J. Osterman, and R. W. Aldrich, "Mutational scanning reveals the determinants of protein stability and aggregation propensity," Structure, vol. 28, no. 1, pp. 125–136.e6, 2020, doi: 10.1016/j.str.2019.10.015.
25. [25] J. Jumper and D. Hassabis, "Protein structure predictions to atomic accuracy with AlphaFold," Nature Methods, vol. 18, no. 10, pp. 1168–1169, 2021, doi: 10.1038/s41592-021-01252-8.
26. [26] J. Xu, "Distance-based protein folding powered by deep learning," Proceedings of the National Academy of Sciences, vol. 116, no. 34, pp. 16856–16865, 2019, doi: 10.1073/pnas.1821309116.
27. [27] A. Roy, A. Kucukural, and Y. Zhang, "I-TASSER: A unified platform for automated protein structure and function prediction," Nature Protocols, vol. 5, no. 4, pp. 725–738, 2010, doi: 10.1038/nprot.2010.5.
28. [28] Y. Zhang, "Progress and challenges in protein structure prediction," Current Opinion in Structural Biology, vol. 18, no. 3, pp. 342–348, 2008, doi: 10.1016/j.sbi.2008.02.004.
29. [29] T. Saito and H. Matsuda, "Protein structure prediction by deep learning representation," BMC Bioinformatics, vol. 22, no. 1, p. 89, 2021, doi: 10.1186/s12859-021-04015-2.
30. [30] S. Wang, J. Peng, J. Ma, and J. Xu, "Folding membrane proteins by deep transfer learning," Cell Systems, vol. 5, no. 3, pp. 202–211.e3, 2017, doi: 10.1016/j.cels.2017.09.001.
31. [31] S. A. Hollingsworth and R. O. Dror, "Molecular dynamics simulation for all," Neuron, vol. 99, no. 6, pp. 1129–1143, 2018, doi: 10.1016/j.neuron.2018.08.011.
32. [32] D. E. Shaw et al., "Atomic-level characterization of the structural dynamics of proteins," Science, vol. 330, no. 6002, pp. 341–346, 2010, doi: 10.1126/science.1187409.
33. [33] D. A. Rufa, J. R. Wagner, and M. Levitt, "Towards chemical accuracy for proteins in molecular dynamics simulations," bioRxiv, 2020, doi: 10.1101/2020.07.29.227959.
34. [34] Y. Hasin, M. Seldin, and A. Lusis, "Multi-omics approaches to disease," Genome Biology, vol. 18, no. 1, p. 83, 2017, doi: 10.1186/s13059-017-1215-1.
35. [35] I. Subramanian, S. Verma, S. Kumar, A. Jere, and K. Anamika, "Multi-omics data integration, interpretation, and its application," Bioinformatics and Biology Insights, vol. 14, 2020, doi: 10.1177/1177932219899051.
36. [36] M. G. S. Costa, D. E. V. S. Rocha, and P. G. Pascutti, "Machine learning in drug discovery and development: Recent progress and future directions," Drug Discovery Today, vol. 26, no. 1, pp. 80–93, 2021, doi: 10.1016/j.drudis.2020.10.010.
37. [37] S. Khan, M. A. Khan, and S. Ahmad, "Computational approaches for predicting the effects of mutations on protein stability: A review," Journal of Bioinformatics and Computational Biology, vol. 19, no. 1, p. 2130001, 2021, doi: 10.1142/S0219720021300013.
38. [38] J. C. Pereira and K. Hamacher, "Using artificial neural networks to reduce computational effort in protein stability predictions," BMC Bioinformatics, vol. 22, no. 1, p. 197, 2021, doi: 10.1186/s12859-021-04119-x.
39. [39] Y. Zhou, Z. Wang, and Y. Yang, "Trends in computational approaches for predicting functionally deleterious single nucleotide variants," Genomics, Proteomics & Bioinformatics, vol. 19, no. 1, pp. 52–60, 2021, doi: 10.1016/j.gpb.2020.06.008.
40. [40] R. Babbush et al., "Quantum computation of electronic transitions using a variational quantum eigensolver," Physical Review X, vol. 8, no. 1, p. 011044, 2018, doi: 10.1103/PhysRevX.8.011044.
41. [41] M. A. Çifçi, P. J. Canatalay, E. Arslan, and S. B. Kausar, "Biyoinspirasyon tabanlı derin öğrenme algoritması," Gazi Üniversitesi Mühendislik Mimarlık Fakültesi Dergisi, vol. 40, no. 2, pp. 979–994, 2023, doi: 10.17341/gazimmfd.1036599.
42. [42] N. Moll et al., "Quantum optimization using variational algorithms on near-term quantum devices," Quantum Science and Technology, vol. 3, no. 3, p. 030503, 2018, doi: 10.1088/2058-9565/aab822.
43. [43] H. M. Berman et al., "The Protein Data Bank at 50: Advancing and sustaining a global data resource," Journal of Biological Chemistry, vol. 296, p. 100743, 2021, doi: 10.1016/j.jbc.2021.100743.
44. [44] S. K. Burley et al., "Protein Data Bank (PDB): The single global macromolecular structure archive," Methods in Molecular Biology, vol. 2112, pp. 3–25, 2020, doi: 10.1007/978-1-0716-0270-6_1.
45. [45] M. Baek et al., "Accurate prediction of protein structures and interactions using a three-track neural network," Science, vol. 373, no. 6557, pp. 871–876, 2021, doi: 10.1126/science.abj8754.
46. [46] A. David, M. J. E. Sternberg, and J. Lees, "Structure, function and disease implications of the 3D genome," Nature Reviews Genetics, vol. 23, no. 2, pp. 101–116, 2022, doi: 10.1038/s41576-021-00408-7.
47. [47] P. Cramer, "AlphaFold2 and the future of structural biology," Nature Structural & Molecular Biology, vol. 28, no. 9, pp. 704–705, 2021, doi: 10.1038/s41594-021-00650-2.
48. [48] M. Hussain et al., "Machine learning-driven approach for a COVID-19 warning system," Electronics, vol. 11, no. 23, p. 3875, 2022, doi: 10.3390/electronics11233875.
49. [49] A. David, M. J. E. Sternberg, and J. Lees, "Structure, function and disease implications of the 3D genome," Nature Reviews Genetics, vol. 23, no. 2, pp. 101–116, 2022, doi: 10.1038/s41576-021-00408-7.
50. [50] U. Akbulut, M. A. Cifci, and Z. Aslan, "Hybrid modeling for stream flow estimation: Integrating machine learning and federated learning," Applied Sciences, vol. 13, no. 18, p. 10203, 2023, doi: 10.3390/app131810203.
51. [51] S. Üzülmez and M. A. Çifçi, "Early diagnosis of lung cancer using deep learning and uncertainty measures," Journal of the Faculty of Engineering and Architecture of Gazi University, vol. 39, no. 1, pp. 385–400, 2024, doi: 10.17341/gazimmfd.1094154.
52. [52] M. Baek et al., "Accurate prediction of protein structures and interactions using a three-track neural network," Science, vol. 373, no. 6557, pp. 871–876, 2021, doi: 10.1126/science. abj8754.
53. [53] D. Wu, S. Liu, H. Moayedi, M. A. Cifci, and B. N. Le, "ANN-incorporated satin bowerbird optimizer for predicting uniaxial compressive strength of concrete," Steel and Composite Structures, An International Journal, vol. 45, no. 2, pp. 281–291, 2022.
54. [54] M. Hussain, A. Islam, J. A. Turi, S. Nabi, M. Hamdi, H. Hamam, and T. Sehar, "Machine learning-driven approach for a COVID-19 warning system," Electronics, vol. 11, no. 23, p. 3875, 2022.