Software and artificial intelligence in the development of modern medicines.
Development of new drugs is a long-term and expensive process. According to J. A. DiMasi et al, the development and implementation of a new drug on average takes about 8-10 years, and the cost of such a project is about 2.8 million US. dollars In Russia, development expenses are about 150-200 million rubles. In carrying out such large-scale tasks, of course, it is necessary to minimize possible risks, as well as to do everything possible to carry out operational work. Specialized software and artificial intelligence (AI) can help with that.
Advances in the field of modern information technologies allow, for example, to assume the possible types of activity of a drug, and, as a consequence, to predict the effectiveness of using the compound for the development of the drug. For example, the PASS-program (Prediction of Activity Spectra for Substances), developed in Russia, allows to make predictions based on structure-activity analysis. The result of the forecast is the probability of any effect (Pa). The selection criterion is generally a probability of Pa ≥ 50 or 70%. In addition, this software allows to determine the probability of occurrence of various adverse reactions.
Most often, PASS is used to select potentially effective and safe chemical compounds among thousands of newly synthesized compounds. For example, in one study of potential substances with anxiolytic effect, 5494 structures were analyzed. According to the screening results, only 8 were selected in the PASS-program for further experiments, which significantly reduced the labor required to select the necessary compounds.
Since 2017, many pharmaceutical companies have also begun to use the services of startups specializing in artificial intelligence. Not so long ago, the international company Insilico Medicine undertook to develop a promising drug for the treatment of complications of breast cancer with the help of artificial intelligence. The main point of the company is analysis of databases with the help of AI to solve certain tasks, for example, searching for new drugs. Also, the startup specializes in the application of AI technologies to increase the productive period of human life.
|Picture 1. Using of AI for the drug discovery. (Reference: Alex Zhavoronkov. Artificial Intelligence for Drug Discovery, Biomarker Development, and Generation of Novel Chemistry.)|
The project is implemented by order of the Sino-Thai research pharmaceutical company Jiangsu Chia Tai FengHai Pharmaceutical, which already has drugs for the treatment of endocrine, respiratory tract, gastrointestinal tract, heart, neurological and oncological diseases in its portfolio system. The cost of the project is about $200 million.
- Alex Zhavoronkov. Artificial Intelligence for Drug Discovery, Biomarker Development, and Generation of Novel Chemistry Molecular Pharmaceutics 2018 15 (10), 4311-4313 DOI: 10.1021/acs.molpharmaceut.8b00930
- Athina Geronikaki et al. Design, synthesis, computational and biological evaluation of new anxiolytics, Bioorganic & Medicinal Chemistry, Volume 12, Issue 24, 2004, Pages 6559-6568, ISSN 0968-0896, https://doi.org/10.1016/j.bmc.2004.09.016.
- Bezhentsev, V.M. & Druzhilovskii, D.S. & Ivanov, Sergey & Filimonov, D.A. & Sastry, G Narahari & Poroikov, Vladimir. (2017). Web Resources for Discovery and Development of New Medicines. Pharmaceutical Chemistry Journal. 51. 1-9. 10.1007/s11094-017-1563-x.
- A. DiMasi, H. G. Grabowski, and R. W. Hansen, J. HealthEcon., 47, 20 – 33 (2016).