Deep Learning-Based Prediction Model for Drug-Target Interactions

  • Authors

    • Puspanjali Mohapatra Assistant professor
    • Shantilata Palei IIIT Bhubaneswar
    • Aryan Sinha IIIT Bhubaneswar
    • P. Subham IIIT Bhubaneswar
    • Adarsh Kumar IIIT Bhubaneswar
    2024-09-22
    https://doi.org/10.14419/3jdke340
  • Drug-target interaction, transfer learning, drug discovery, deep learning, SMILES, CNN, RNN, Transformer.

  • Abstract

    This paper comprehensively studies deep learning approaches for drug-target interaction (DTI) prediction in drug discovery. We evaluate the performance of convolutional neural networks (CNNs), recurrent neural networks (RNNs), and transformers on DTI prediction tasks. Our results demonstrate that the CNN model consistently outperforms both RNN and transformer models in accuracy. Additionally, we investigate the impact of transfer learning on DTI model performance, showing that pre-trained fine-tuning significantly enhances the results. These insights contribute to selecting and optimising deep learning models for DTI prediction, thereby advancing drug discovery efforts. Notably, our findings highlight the potential of combining CNNs with the BindingDB dataset and utilizing transformers as pretrained models for real-world DTI cases.

  • References

    1. Dongmin Bang, Bonil Koo, and Sun Kim. Transfer learning of condition-specific perturbation in gene interactions improves drug response prediction.
    2. Bioinformatics, 40(Supplement 1):i130–i139, 2024.
    3. Chenjing Cai, Shiwei Wang, Youjun Xu, Weilin Zhang, Ke Tang, Qi Ouyang, Luhua Lai, and Jianfeng Pei. Transfer learning for drug discovery. Journal
    4. of Medicinal Chemistry, 63(16):8683–8694, 2020.
    5. Ruolan Chen, Xiangrong Liu, Shuting Jin, Jiawei Lin, and Juan Liu. Machine learning for drug-target interaction prediction. Molecules, 23(9):2208,
    6. Yanyi Chu, Xiaoqi Shan, Tianhang Chen, Mingming Jiang, Yanjing Wang, Qiankun Wang, Dennis Russell Salahub, Yi Xiong, and Dong-Qing Wei.
    7. Dti-mlcd: predicting drug-target interactions using multi-label learning with community detection method. Briefings in bioinformatics, 22(3):bbaa205,
    8. Ali Ezzat. Challenges and solutions in drug-target interaction prediction. PhD thesis, 2018.
    9. Wenbo Guo, Yawen Dong, and Ge-Fei Hao. Transfer learning empowers accurate pharmacokinetics prediction of small samples. Drug Discovery
    10. Today, page 103946, 2024.
    11. Yi-Sue Jung, Yoonbee Kim, and Young-Rae Cho. Comparative analysis of network-based approaches and machine learning algorithms for predicting
    12. drug-target interactions. Methods, 198:19–31, 2022.
    13. Mohammad Reza Keyvanpour, Faraneh Haddadi, and Soheila Mehrmolaei. Dtip-tc2a: An analytical framework for drug-target interactions prediction
    14. methods. Computational Biology and Chemistry, 99:107707, 2022.
    15. Ashish Khanna, Deepak Gupta, Siddhartha Bhattacharyya, Aboul Ella Hassanien, Sameer Anand, and Ajay Jaiswal. International conference on
    16. innovative computing and communications. Proceedings of ICICC, 2, 2021.
    17. Ingoo Lee, Jongsoo Keum, and Hojung Nam. Deepconv-dti: Prediction of drug-target interactions via deep learning with convolution on protein
    18. sequences. PLoS computational biology, 15(6):e1007129, 2019.
    19. Bin Liu. Addressing class imbalance in multi-label data. PhD thesis, Aristotle University Of Thessaloniki, Greece, 2021.
    20. Seyedeh Zahra Sajadi, Mohammad Ali Zare Chahooki, Sajjad Gharaghani, and Karim Abbasi. Autodti++: deep unsupervised learning for dti prediction
    21. by autoencoders. BMC bioinformatics, 22:1–19, 2021.
    22. Wen Shi, Hong Yang, Linhai Xie, Xiao-Xia Yin, and Yanchun Zhang. A review of machine learning-based methods for predicting drug–target
    23. interactions. Health Information Science and Systems, 12(1):1–16, 2024.
    24. Derwin Suhartono, Muhammad Rizki Nur Majiid, Alif Tri Handoyo, Pandu Wicaksono, and Henry Lucky. Towards a more general drug target
    25. interaction prediction model using transfer learning. Procedia Computer Science, 216:370–376, 2023.
    26. AA Toropov, AP Toropova, and E Benfenati. Qsar-modeling of toxicity of organometallic compounds by means of the balance of correlations for
    27. inchi-based optimal descriptors. Molecular diversity, 14:183–192, 2010.
    28. Orhan Gazi Yalc¸ın and T Istanbul. Applied neural networks with TensorFlow 2: API oriented deep learning with python. Springer, 2021.

  • Downloads

  • How to Cite

    Mohapatra, P., Palei, S., Sinha, A., P. Subham, & Adarsh Kumar. (2024). Deep Learning-Based Prediction Model for Drug-Target Interactions. Journal of Advanced Computer Science & Technology (JACST), 12(2), 53-59. https://doi.org/10.14419/3jdke340