Computational Prediction of the Comprehensive SARS-CoV-2 vs. Human Interactome to Guide the Design of Therapeutics

Kevin Dick; Kyle K Biggar; James R Green

The highlighted sites may appear "shifted" relative to the highlighted cells (typically in the bottom-left); this is due to the algorithm's use of a window of 20 amino acids in length that extends both to the left (along the x-axis) and upwards (along the y-axis). Consequently, the minimum PIPE-Site size is 20×20 with the peak in the bottom-left corner. Additionally, this implementation may result in the predicted site extending past the coloured matrix, either to the right or above. This defined window size additionally prevents predictions within the terminal 20 amino acids of both sequences given that the widow sizes in these regions would necessarily be less than 20 amino acids in length. Finally, the PIPE-Sites may overlap when numerous hits appear within close proximity, as is the case when a "band" of hits appears in the matrix. Finally, when the peak of the landscape comprises only a few hits (generally <3) the entire landscape is predicted as a site of interaction; evidently, these should be disregarded. Therefore, when interpreting the landscapes, it is important not to solely rely on these proposed regions; they function as an initial guide, yet other high-scoring, or "hot-spot", regions of interest may exist in the landscape. By providing a matrix of raw scores (in the form of a space-separated .mat file), visual interpretation of the results promise to reveal notable subsequences as well as enable the application of related interaction site predictors to identify putative sites of interaction. The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/2020.03.29.014381 doi: bioRxiv preprint Figure 6 : The PIPE-Sites Landscape between the SARS-CoV-2 Spike protein and human ACE2 protein. The three red rectangles represent the predicted PIPE-Sites regions.