17th International Conference on Fundamental and Applied Aspects of Physical Chemistry (Proceedings, Volume II) (2024) [O-04-O, pp. 599-602]
AUTHOR(S) / AUTOR(I): Branislav Stanković 
Download Full Pdf 
DOI: 10.46793/Phys.Chem24II.599S
ABSTRACT / SAŽETAK:
A novel procedure for descriptor selection was tested on a quantitative structure–activity relationship (QSAR) model for SARS-CoV-2 main protease inhibitors, aimed at developing transparent, interpretable, reproducible, and publicly available methodologies applicable in drug development and other QSAR-related fields. The model was trained and tested using molecules collected from the CHEMBL database. Descriptors were selected from a set of Mordred molecular descriptors. The models were constructed using multiple linear regression algorithms, followed by scrutiny of fitting and predictive performance, reliability, and robustness through various statistical validation criteria. Among the models, those trained with ordinary least squares linear regression exhibited the best performances. Due to their good predictive performances the constructed models can serve as valuable tools for the quick and reliable prediction of inhibitory activity toward SARS- CoV-2 main protease.
KEYWORDS / KLJUČNE REČI:
ACKNOWLEDGEMENT / PROJEKAT:
This work was supported by the Ministry of Education, Science and Technological Development of the Republic of Serbia (grant no No. 451-03-66/2024-03/200017).
REFERENCES / LITERATURA:
- P. Zhou, X. L. Yang, X. G. Wang, B. Hu, L. Zhang, W. Zhang, H. R. Si, Y. Zhu, B. Li, C. L. Huang, Nature, 579 (2020) 270.
- https://data.who.int/dashboards/covid19.
- T. Zhai, F. Zhang, S. Haider, D. Kraut, Z. Huang, Protease. Front. Mol. Biosci., 8 (2021) 661424.
- Z. Jin, X. Du, Y. Xu, et al.., Nature. 582 (2020) 289.