Generating protein folding trajectories Using contact-map-driven directed walks

With the advent of AlphaFold, predicting the folded structure of any given protein has become much easier. And yet, AlphaFold doesn’t tell you anything about how exactly the protein folds: what intermediate states, if any, does it visit before folding; what about misfolded structures; and what is the kinetic of the folding process? To answer any of these questions, we need to take a step back and think of clever ways in which we can avoid the so-called timescale problem – i.e., the fact that protein folding usually happens on timescale far beyond the reach of what we can simulate by, for instance, classical molecular dynamics. This work, led by Scott Habershon (the oldest and wisest member of Warwick Catch) and Ziad Fakhoury (a PhD student within the HetSys CDT), puts forward an ingenious methodology that leverages “simple” graphs to gain insight into the folding mechanism. This is the start of a long road – here, we used a simple model protein, but would we be able to apply this to big proteins as well? Stay tuned! Meanwhile, you can have a look at the paper online or grab a pdf here.