De novo 3D models of SARS-CoV-2 RNA elements and small-molecule-binding RNAs to guide drug discovery

Ramya Rangan; Andrew M. Watkins; Wipapat Kladwang; Rhiju Das

The modeling also included riboswitch aptamer cases (De novo, Figs. 5D-J) in which the ligand binding sites were not modeled by homology, in closer analogy to virtual screening approaches that might make use of the FARFAR2-SARS-CoV-2 models. These models include cases such as the ydaO riboswitch, where modeling did not achieve a model closer than 10.0 Ã… RMSD to the RNA . CC-BY 4.0 International license author/funder. It is made available under a The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/2020.04.14.041962 doi: bioRxiv preprint crystallized with two cyclic-diAMP ligands, perhaps owing to the large ligand and the substantial degree to which contacts with that ligand may organize the crystallized conformation. It will be interesting to see if these models still allow recognition of small molecule binding sites by computational methods. We provide in Table 2 additional metrics for these model sets, including the same RMSD convergence estimates, cluster occupancy, and E-gap numbers as for our FARFAR2-SARS-CoV-2 models as well as RMSD to experimentally determined ligand-bound structures. It will be interesting to see if these metrics give any correlation with the ability of virtual screening methods to discover known native ligands for these RNA elements, which will be useful in evaluating the likelihood of success of such methods in discovering molecules for the FARFAR2-SARS-CoV-2 models.