Grammatical evolution-based design of nucleotic analogs for SARS-CoV-2's replication-transcription complex.

In the context of de novo structure based drug design, the rise of in silico tools shows a promising strategy to search new molecules to address several diseases. In this work, a program with a Grammatical Evolution approach, lead to the design of 4 new molecules that can inhibit the SARS-CoV-2 RNA dependent RNA polymerase. Those were build by an exploration of a chemical space, defined previously by a pharmacophore model, with several bioisosteric fragments arrange by the grammatical evolution code into 1D molecular strings, which were then converted into 3D molecules and evaluated against the desire target using automated docking calculations. The novel drug candidates demonstrate the capacity to manifest ligand efficiencies at the biological target that are analogous to those exhibited by Remdesivir. The molecules in question also demonstrate pharmacokinetic profiles that are analogous. Two of these molecules have been observed to maintain stable interaction properties with the RNA-dependent RNA polymerase over the course of 200 nanoseconds of molecular dynamics simulation. These molecules primarily interact with the active site through weak hydrogen bonds. The software we developed for this contribution is available in the following URL: https://github.com/masotelof/GEMolecularDesign.