The Use of In Silico Tools for the Toxicity Prediction of Potential Inhibitors of SARS-CoV-2

Abstract

The current strategy for treating the Covid-19 coronavirus disease involves the repurposing of existing drugs or the use of convalescent plasma therapy, as no specific therapeutic intervention has yet received regulatory approval. However, severe adverse effects have been reported for some of these repurposed drugs. Recently, several in silico studies have identified compounds that are potential inhibitors of the main protease (3-chymotrypsin-like cysteine protease) and the nucleocapsid protein of SARS-CoV-2. An essential step of drug development is the careful evaluation of toxicity, which has a range of associated financial, temporal and ethical limitations. In this study, a number of in silico tools were used to predict the toxicity of 19 experimental compounds. A range of web-based servers and applications were used to predict hepatotoxicity, mutagenicity, acute oral toxicity, carcinogenicity, cardiotoxicity, and other potential adverse effects. The compounds were assessed based on the consensus of results, and were labelled as positive or negative for a particular toxicity endpoint. The compounds were then categorised into three classes, according to their predicted toxicity. Ten compounds (52.6%) were predicted to be non-mutagenic and non-hERG inhibitors, and exhibited zero or low level hepatotoxicity and carcinogenicity. Furthermore, from the consensus of results, all 19 compounds were predicted to be non-mutagenic and negative for acute oral toxicity. Overall, most of the compounds displayed encouraging toxicity profiles. These results can assist further lead optimisation studies and drug development efforts to combat Covid-19.

Natural Phytocompounds from Common Indian Spices for Identification of Three Potential Inhibitors of Breast Cancer: A Molecular Modelling Approach

“…At the same time, the Protox II webserver ( accessed on 21 June 2022) was used to predict the toxicity of the ligands [ 54 ]. ProTox-II predicts values of various toxicity endpoints such as hepatotoxicity, carcinogenicity, cytotoxicity, mutagenicity, immunotoxicity, along with a predicted median lethal dose (LD 50 ) and toxicity class [ 54 , 55 ].…”

Synthesis, Characterization and Antimicrobial Evaluation for New Esters Derivatives Containing Two 1, 3, 4-Oxadiazole Units

“…To predict the toxicity of the prepared compounds, the following program was used: Pro-Tox-II: Prediction of Toxicities of Chemicals Online https://tox-new.charite.de, , [20,21,22], and the prediction of the toxicity of a synthetic compound with rats and the determination of LD50 online http://www.way2drug.com/passonline, [23,24]…”

Artificial Intelligence in Drug Toxicity Prediction: Recent Advances, Challenges, and Future Perspectives

“…Previously, the rodent acute oral toxicity test was predicated on the calculation of LD 50 . However, owing to the limitations arising from interspecies variability and various ethical concerns, new testing methodologies have supplanted classic LD 50 testing . Alternative methods are being explored to replace acute toxicity assays in animals, including in vitro (tissue culture) and in silico assessment and prediction of LD 50 …”

“…However, owing to the limitations arising from interspecies variability and various ethical concerns, new testing methodologies have supplanted classic LD 50 testing. 43 Alternative methods are being explored to replace acute toxicity assays in animals, including in vitro (tissue culture) and in silico assessment and prediction of LD 50 . 44 Recently, two new software tools for ADMET prediction have become available, namely, FP-ADMET 15 and Interpretable-ADMET, 29 both of which have built-in LD 50 prediction models.…”