Folding@home

Folding@home (FAH) is a distributed computing project focused on simulating the physical process by which proteins fold from a linear chain of amino acids into their functional three-dimensional structures. Protein misfolding is implicated in many serious diseases, including Alzheimer's, Parkinson's, Huntington's, ALS, and many cancers. By simulating these molecular processes at atomic resolution, researchers gain insights that are impossible to obtain through laboratory experiments alone.

The project was launched in October 2000 at Stanford University by Professor Vijay Pande. Its core scientific method uses molecular dynamics (MD) simulations — numerically solving Newton's equations of motion for every atom in a protein system, typically tens of thousands of atoms interacting through complex force fields. Each simulation step covers only femtoseconds (10^-15 seconds), meaning that simulating biologically relevant timescales of microseconds to milliseconds requires an astronomical number of computational steps.

FAH pioneered the Markov State Model (MSM) approach, which breaks long simulations into many short, independent trajectories that can be run in parallel on volunteers' computers. These short trajectories are then statistically assembled into a complete picture of the protein's folding landscape — the set of all possible conformations and the transition rates between them. This approach is ideally suited to distributed computing, because each work unit is independent and relatively small.

During the COVID-19 pandemic in early 2020, FAH launched an emergency project to simulate the SARS-CoV-2 spike protein and its receptor binding domain. Volunteers flooded the project — within weeks, FAH reached a combined computing speed of approximately 2.4 exaFLOPS, surpassing the combined power of the world's top 500 supercomputers. The simulations revealed "cryptic" binding pockets on the spike protein that were not visible in static crystal structures, providing potential targets for therapeutic intervention.

To date, FAH has produced over 280 peer-reviewed scientific publications. In 2019, Professor Greg Bowman succeeded Vijay Pande as director while based at Washington University in St. Louis; he brought the project with him to the University of Pennsylvania in 2022, where it remains headquartered today. The project's long history of simulating protein dynamics contributed foundational data and methods that underpin the computational protein design work recognized by the 2024 Nobel Prize in Chemistry.