Оцінка цитотоксичності аліфатичних амінокарбонових сполук як потенційних протикоронавірусних агентів

Background. Despite the success in creating vaccines against SARS-CoV-2, the high mutagenicity of coronaviruses, interspecies transmission, and the emergence of new strains require further search for effective antiviral agents. A key step in this process is to evaluate the cytotoxicity of potential compounds to determine their safety and therapeutic potential. Modern IT solutions, such as automated image analysis and artificial intelligence, increase the accuracy and objectivity of assessments. Objective. To determine the cytotoxicity of compounds with potential anticoronavirus activity and to analyze it using IT tools. Methods. The study used the grafting cell line BHK-21 of the gerbil hamster, which was incubated with seven aliphatic amino carbon compounds in six concentrations. Cell viability was determined using the MTT assay. Cell monolayer image processing and an exponential dose-response model were used for automated analysis. Results. The study revealed a pronounced dose- and time-dependent cytotoxicity of most samples, with a maximum decrease in viability at concentrations above 10 mg/ml. The hormesis effect was recorded at low concentrations (up to 5-10 mg/ml), which may indicate the activation of cellular defense mechanisms. The high correlation between measurements at 492 nm and 550 nm (R² > 0.98) confirmed the reliability of the spectrophotometric data. The exponential model allowed us to approximate the toxicity curves, especially in the middle and high concentration ranges. The built neural network based on image data and MTT test showed the ability to predict cell viability even with a limited amount of training data. Conclusions. The combination of the MTT assay with automated image analysis provides a comprehensive assessment of cytotoxicity. A dose-dependent decrease in cell viability and morphological changes under the influence of the studied compounds were found. Measurements at 550 nm proved to be more sensitive to early changes in cell metabolism. The use of IT algorithms has demonstrated the prospects of an automated approach to the screening of biologically active substances.