We are hosting research seminars of topical interest to the biosimulation community. These are open to attend for anyone with a keen interest in the life sciences and molecular simulation, we ask that you register to each event, so that you receive the links to the talks.
Prediction of lipid contacting residues based on the simulation data of membrane proteins
26th of April 2021 at 11:00 until 12:00 - registration open now
Speaker: Chen Song Assistant Professor, Center for Quantitative Biology, Peking University, China
Membrane proteins are essential for transmembrane signaling and transport. We use multi-scale molecular dynamics (MD) simulations to investigate the function mechanisms of these transmembrane machinery. With more and more physics-based simulation data available, we are also trying to use machine learning-based methods to make more efficient predictions on the structural and functional properties of membrane proteins. In this talk, I will present our recent effort using machine learning to predict which amino acids of a protein are likely to be in direct contact with the hydrophobic core of membranes, based on the MemProtMD, a highly informative MD-generated database developed by Phillip Stansfeld and Mark Sansom et al. The lipid-contacting information can help to improve the prediction precision of protein contact maps, and perhaps can also be used to screen membrane-anchored proteins from sequences.