Binding site analysis of potential inhibitors of SARS-CoV-2 S2
Therapeutics are needed to control the spread of SARS-CoV-2 and the S2 region is an effective target due to its highly conserved nature. In this project, we will find examples of existing inhibitors targeting this region in similar viruses (i.e. SARS-CoV, MERS-CoV) and dock them to SARS-CoV-2 S2 to understand the binding site interactions. Then, we will re-design these inhibitors to best fit SARS-CoV-2 S2 and improve upon our binding interactions in the final project.
Application of cryoDRGN for cryo-EM image analysis
Cryo-EM is a revolutionary imaging technique to capture the molecule representation of proteins, though AI is required to denoise and reconstruct the 2D projections of these proteins. CryoDRGN is a recent paper demonstrating improved ability to reconstruct proteins through neural networks, so in this project we will validate the results of cryoDRGN by applying it to another dataset from the EMPIRE database as well as learn the details of neural networks and cryo-EM image analysis.
Gene network analysis in human disease
Gene networks are promising methods to predict connections between genes and human disease. In this project, we will use gene expression and biology pathways (from Enrichr) to develop a neural network to predict disease (i.e. COVID-19).