Folding @ home

from Wikipedia, the free encyclopedia
Folding @ home
F @ H Logo 2012.png
Area: Medicine and biology
Target: Simulation of protein folding
Operator: Stanford University
Country: United StatesUnited States United States
Platform: Windows , Linux , macOS , FreeBSD formerly: BOINC , Android
Website: foldingathome.org
Project status
Status: active
Start: September 19, 2000
The End: active

Folding @ home (often also F @ H or FAH for short ) is a volunteer computing project for disease research that simulates protein folding and other types of molecular dynamics . Instead of using the computing power of a single computer, a complex task is divided into sub-tasks, these are distributed over several computers and their computing power is used to complete the task. Through distributed computing, the project uses the unused processing resources of personal computers and servers on which the software is installed and which thus contribute to research into diseases.

The project uses a statistical simulation method that represents a paradigm shift compared to traditional calculation methods . As part of the client-server model , after requesting a service from the server, the participants (clients) receive parts of a simulation (work units), calculate and complete them and return them to the project's database server, where the units are then can be compiled into an overall simulation.

The main purpose of the project is to determine the mechanisms of protein folding, i.e. H. the process by which proteins achieve their final three-dimensional structure and the investigation of the causes of protein misfolding. This is of interest for medical research on Alzheimer's , Huntington's, and many forms of cancer, among other diseases. To a lesser extent, Folding @ home also seeks to predict the ultimate structure of a protein and determine how other molecules can interact with it, which affects drug development .

On April 13, 2020, during the COVID-19 pandemic, Folding @ Home achieved combined computing power faster than the 500 fastest supercomputers in the world put together, outperforming the fastest supercomputer at the time by 15 times.

On April 17, 2020, new client software was released that adds COVID-19 projects to the list of prioritizable projects.

history

According to Vijay Pande, the first ideas for using numerous computers came up in the summer of 2000, when the first version of the client software was programmed - at that time still a screensaver . On September 19, 2000, the first software for Folding @ home was officially released by the Pande Laboratory at Stanford University and has since been developed on a non-profit basis under the direction of Vijay Pande . It has been under the direction of Gregory Bowman , Professor of Biochemistry and Molecular Biophysics at Washington University School of Medicine in St. Louis , since 2019 - it is used collectively by various scientific institutions and research laboratories around the world. The results generated by Folding @ home are not sold. Researchers worldwide can call up the generated data sets on request and obtain them directly from a website.

Folding @ home researcher Gregory Bowman received the 2010 Thomas Kuhn Paradigm Shift Award from the American Chemical Society for developing the open source software MSMBuilder and for achieving quantitative agreement between theory and experiment. For his work, Vijay Pande was again awarded the Michael and Kate Bárány Prize for Young Researchers in 2012 for the “development of computational methods for creating leading theoretical models for protein and RNA folding” and in 2006 with the Irving Sigal Young Investigator Award for his Simulation results excellent.

background

A protein before and after its folding. The protein is initially available as a random coil and is then folded in such a way that it is in its native conformation .

Proteins are an essential part of many biological functions and are involved in all processes that take place in cells. Often such proteins are enzymes that carry out biochemical reactions, including signal transduction , molecular transport, and cell regulation. Some proteins act as structural proteins , which act as a kind of scaffold for cells, while other proteins, like antibodies , are involved in the immune system . In order for a protein to perform these functions, it must fold into a functional three-dimensional structure, a process that is often spontaneous and depends on interactions within its amino acid sequence and interactions between the amino acids and its environment.

Protein folding is mainly determined by hydrophobic interactions , the formation of intramolecular hydrogen bonds, and van der Waals forces that counteract conformational entropy. The folding process often begins co-translationally , so that the N -terminal of the protein begins to fold while the C -terminal part of the protein is still being synthesized by the ribosome . However, a protein molecule can fold spontaneously during or after biosynthesis. The folding process also depends on the solvent (water or double lipid layer ), the salt concentration, the pH , the temperature, the possible presence of cofactors and molecular chaperones . Proteins have limitations with regard to their folding possibilities due to steric hindrances between individual atoms, so that only certain combinations of dihedral angles are allowed. These allowable angles of protein folding are described in a two-dimensional diagram known as the Ramachandran diagram, described with φ ( Phi ) and ψ ( Psi ) angles.

Understanding protein folding is important for bioinformatics in order to predict certain functions and mechanisms of the protein. Although the folding takes place in a cellular environment that has a high concentration of proteins ( known as macromolecular crowding ), it usually goes smoothly. Due to the chemical properties of a protein or other factors, misfolding of the protein can occur. Even if cellular mechanisms contribute to misfolded proteins being removed or refolded, the misfolded proteins can aggregate and cause a variety of diseases. Laboratory experiments to investigate protein folding processes are limited in terms of their scope and accuracy, which is why physical calculation models are used that provide a deeper understanding of protein folding, misfolding and aggregation.

Due to the complexity of the protein conformation or the configuration space of the protein (the set of all possible folding states that a protein can assume) and the limited computing power, molecular dynamics simulations are severely limited with regard to the investigation of protein folds. While most proteins typically fold on the order of milliseconds, before 2010 simulations could only reach timescales from nanoseconds to microseconds. Supercomputers have been used to simulate protein folding, but such systems are costly and, for the most part, are shared by many research groups. Since the calculations in kinetic models take place one after the other, scaling traditional molecular dynamics simulations to such systems is extremely difficult. In addition, since protein folding is a stochastic process and can vary statistically over time, it is computationally difficult to use long simulations to get comprehensive views of the folding process.

Folding @ home uses Markov state models to model the possible shapes and folding pathways a protein can adopt when it condenses from its initial random-coil state (left) to its native 3-D structure (right).

The protein folding does not take place in one step. Instead, proteins spend most of the total folding time, in some cases up to 96%, in various intermediate conformational states, each of which represents a local thermodynamic minimum of free energy in the protein's energy landscape (see folding funnel ). Through a process called Adaptive Sampling , these conformations are used by Folding @ home as starting points for a series of simulation runs for folding processes. Over time, new conformations are discovered that serve as new starting points for simulation processes (cyclic process). As soon as one can assign a transmission probability to the underlying hidden states and an emission probability to the externally observable output symbols (so-called emissions) resulting from the hidden states, one speaks of the Hidden Markov Model (HMM). HMM are discrete-time master equation models that describe the conformational and energy landscape of a biomolecule as a set of different structures and short transition states between the structures. The Hidden Markov Model, combined with Adaptive Sampling, increases the efficiency of the simulation considerably. Since this avoids the calculation within the local energy minimum, this combination is suitable for distributed systems (including GPUGRID ) as it enables the statistical accumulation of short, independent simulation processes for convolution processes. The time it takes to create a Hidden Markov Model is inversely proportional to the number of parallel simulations, i.e. H. the number of processors available . In other words: a parallel processing is created (see parallel computer ), which leads to a reduction of the total computation time by approximately four orders of magnitude. A closed Hidden Markov Model can contain up to ten thousand states from the phase space of the protein (all the conformations that a protein can adopt) and the transitions between them. The model illustrates folding processes and paths, and researchers can later use kinetic clusters to create a so-called coarse-grained representation of the detailed model. These hidden Markov models can be used to determine misfolding processes and to compare simulations quantitatively with experiments.

Between 2000 and 2010, the length of an amino acid sequence of the proteins examined by Folding @ home increased by a factor of four, while the time scales for protein folding simulations increased by six orders of magnitude. In 2002, Folding @ home used Markov state models to run a processor time of approximately one million days with simulations over a period of several months, and in 2011 another simulation was processed in parallel, which required a total of 10 million computing hours processor time. In January 2010, Folding @ home used HMM to simulate the dynamics of the slowly folding NTL9 protein with 32 amino acid residues with a simulation time of 1.52 milliseconds, a timescale that agrees with experimental predictions of the folding rate but is a thousand times longer than the time previously achieved . The model consisted of many individual trajectories , each two orders of magnitude shorter, and provided an accurate representation of the protein's energy landscape.

Applications in biomedicine

Viral diseases

Folding @ home supports research to prevent some viruses, such as influenza and HIV , from recognizing and invading biological cells. In 2011, Folding @ home began simulating the dynamics of the enzyme RNase H, a key component of HIV, to try to develop drugs that deactivate this enzyme. Folding @ home has also been used to study membrane fusion , a key event in viral infection and a variety of biological functions. This fusion involves conformational changes of the viral fusion proteins and the docking of the proteins, but the exact molecular mechanisms behind the fusion are largely unknown. Fusion events can consist of over half a million atoms that interact for hundreds of microseconds. The development of models to predict the mechanisms of membrane fusion contributes to the scientific understanding of the process with antiviral drugs . In 2006, scientists used Markov state models and the Folding @ home network to discover two avenues for fusion and gain further insights.

After detailed folding @ home simulations of small cells known as vesicles , the Pande laboratory introduced a new computational method in 2007 to measure the topology of structural changes during fusion. In 2009, researchers used Folding @ home to study mutations in influenza hemagglutinin , a protein that binds a virus to its host cell and helps the virus to enter. Mutations in hemagglutinin affect how well the protein binds to receptor molecules on a host's cell surface, which determines how infectious the virus strain is to the host organism. Knowing the effects of hemagglutinin mutations helps in developing antiviral drugs. Since 2012, Folding @ home has continued to simulate the folding and interactions of hemagglutinin, complementing experimental studies at the University of Virginia .

SARS-CoV-2 virus

SARS-CoV-2 RBD together with human antibodies
Ribbon model of the SARS-CoV-2-M (pro) protease of the SARS-CoV-2 coronavirus as a target for protease inhibitors .

In March 2020, Folding @ home launched a program to support researchers around the world who are working to find a cure and learn more about the outbreak of COVID-19 - also known as the respiratory disease caused by the novel coronavirus - to experience. The first wave of projects simulates potentially drug-treatable protein targets of the SARS-CoV-2 virus and the related SARS-CoV virus, of which there are significantly more data.

Alzheimer's disease

The Alzheimer 's affects an incurable neurodegenerative disorder that primarily the elderly and for more than half of all dementia cases is responsible. The exact cause remains unknown, but the disease is identified as a protein misfolding disease. Alzheimer's is associated with toxic aggregations of the peptide beta-amyloid (Aβ) caused by misfolding and clumping of Aβ together with other Aβ peptides. These Aβ aggregates then grow into significantly larger senile plaques, a pathological marker of Alzheimer's disease. Because of the heterogeneous nature of these aggregates, experimental methods such as X-ray crystallography and nuclear magnetic resonance (NMR) have had difficulty characterizing their structures. In addition, the atomic simulations of the Aβ aggregation are computationally very demanding due to their size and complexity.

Preventing Aβ aggregation is a promising method of developing therapeutics for Alzheimer's disease, according to Doctors Naeem and Fazili in a review article. In 2008, Folding @ home simulated the dynamics of the aggregation of Aβ in atomic detail over time scales on the order of ten seconds. Previous studies could only simulate about 10 microseconds. Folding @ home was able to simulate the folding of Aβ six orders of magnitude longer than previously possible. The researchers used the results of this study, a beta-hairpin ( beta-hairpin to identify), which was a major source of molecular interactions within the structure. The study helped prepare the Pande Laboratory for future aggregation studies and further research to find a small peptide that could stabilize the aggregation process.

In December 2008, Folding @ home found several small drug candidates that appear to inhibit the toxicity of the Aβ aggregates. In 2010, in close cooperation with the Center for Protein Folding Machinery, a start was made on testing these drug candidates on biological tissue. In 2011, Folding @ home completed simulations of several mutations in Aβ that appear to stabilize aggregate formation, which could aid the development of therapeutic drug therapies for the disease and be very useful in experimental nuclear magnetic resonance spectroscopy studies on Aβ oligomers. Later that year, Folding @ home began simulating various Aβ fragments to determine how various natural enzymes affect the structure and folding of Aβ.

Chorea huntington

The Huntington's disease is a neurodegenerative genetic disease associated with misfolding and aggregation associated proteins. Excessive repetitions of glutamic acid at the N-terminus of the huntingtin protein cause aggregation, and while the behavior of the repetitions is not fully understood, it does lead to the cognitive decline associated with the disease. As with other aggregates, there are difficulties in determining their structure experimentally. Scientists are using Folding @ home to study the structure of the huntingtin protein aggregate and predict how it will form, while helping rational drug design to stop aggregate formation. The N17 fragment of the HD protein accelerates this aggregation, and although several mechanisms have been suggested, its exact role in the process is still largely unknown. Folding @ home simulated this and other fragments to clarify their role in the disease. Since 2008, his Alzheimer's disease drug design methods have been applied to Huntington's disease.

cancer

More than half of all known types of cancer are mutations in p53 , a tumor suppressor protein found in every cell that regulates the cell cycle and signals cell death when the DNA is damaged . Specific mutations in p53 can disrupt these functions, so that an abnormal cell can continue to grow in an uncontrolled manner, which leads to the development of tumors . Analysis of these mutations helps explain the root causes of p53-related cancers. In 2004, Folding @ home was the first to conduct the first molecular dynamics study on refolding the p53 protein dimer in a purely atomic simulation of water. The results of the simulation were in agreement with experimental observations and provided insights into the refolding of the dimer that were not previously possible. This was the first peer-reviewed publication on cancer from a distributed computer project. The following year, Folding @ home, a new method of identifying the amino acids that are critical to the stability of a particular protein, was used to study mutations in p53. The method was reasonably successful in identifying cancer-promoting mutations and determined the effects of specific mutations that otherwise could not be measured experimentally.

Folding @ home is also used to study chaperones , heat shock proteins that play an essential role in cell survival by helping other proteins to fold in the crowded and chemically stressful environment within a cell. Rapidly growing cancer cells rely on specific chaperones, and some chaperones play a key role in chemotherapy resistance . The inhibition of these specific chaperones is seen as a potential mode of action for efficient chemotherapeutic agents or to reduce the spread of cancer. With Folding @ home and in close cooperation with the Center for Protein Folding Machinery , the Pande Laboratory hopes to find a drug that inhibits the chaperones involved in cancer cells. Researchers are also using Folding @ home to study other cancer-related molecules , such as the enzyme Src kinase and some forms of the engraved homeodomain : a large protein that could be implicated in many diseases, including cancer. In 2011, Folding @ home began simulating the dynamics of the small protein EETI, which binds to surface receptors on cancer cells to identify carcinomas in imaging processes.

Interleukin 2 (IL-2) is a protein that helps the immune system's T cells to attack pathogens and tumors. However, its use as a cancer treatment is limited because of serious side effects such as pulmonary edema . IL-2 binds to these lung cells differently than it does to T cells, so IL-2 research needs to understand the differences between these binding mechanisms. In 2012, Folding @ home supported the discovery of a mutated form of IL-2 that is three hundred times more effective in its role as an immune system but has fewer side effects. In experiments, this altered form has significantly outperformed the natural IL-2 in preventing tumor growth. Pharmaceutical companies have shown interest in the mutant molecule, and the National Institutes of Health are testing it against a variety of tumor models to accelerate its development as a therapeutic.

Osteogenesis imperfecta

Osteogenesis imperfecta , known as glass bone disease, is an incurable genetic bone disease that can be fatal. The sufferers are unable to form functional connective tissue . This is mostly due to a mutation in type I collagen, which performs a variety of structural functions and is the most abundant protein in mammals. The mutation causes a deformation of the triple helix structure of collagen which, if not naturally destroyed, results in abnormal and weakened bone tissue. In 2005, Folding @ home tested a new quantum mechanical method that improved previous simulation methods and that could be useful for future computer studies of collagen. Although researchers have used Folding @ home to study collagen folding and misfolding, the interest in this project compares to Alzheimer's and HD research as a pilot project.

Drug design

Drugs work by binding to specific sites on the target molecules and causing a desired change, such as: B. deactivating a target or causing a conformational change . Ideally, a drug should have a very specific effect and only bind to its target molecule without impairing other biological functions. However, it is difficult to determine exactly where and how tightly two molecules will bind. Due to the limited computing power, today's methods in silico usually have to trade speed for accuracy - e.g. For example, fast protein docking methods must be used instead of computationally intensive free energy calculations. The computing power of Folding @ home allows researchers to use both methods and evaluate their efficiency and reliability. Computer-aided drug design has the potential to accelerate drug development and reduce costs. In 2010, Folding @ home used MSMs and free energy calculations to predict the native state of the villin protein with a deviation of up to 1.8 Angstroms (Å) via RMSD ( root mean square deviation ) from the experimental crystal structure determined by X-ray crystallography . This accuracy has implications for future methods of predicting protein structure , including for inherently unstructured proteins. Scientists have used Folding @ home to research drug resistance by studying vancomycin , an antibiotic of last resort, and beta-lactamase , a protein that can break down antibiotics like penicillin .

Chemical activity takes place along the active site of a protein. Traditional methods of drug design involve tightly tying this site and blocking its activity, assuming that the target protein exists in a rigid structure. However, this approach only works for about 15% of all proteins. Proteins contain allosteric sites which, when bound by small molecules, can change the conformation of a protein and ultimately affect the activity of the protein. These locations are attractive drug destinations, but are computationally expensive to locate. In 2012, Folding @ home and MSMs were used to identify allosteric sites in three medically relevant proteins: beta-lactamase, interleukin-2 and RNase H.

About half of all known antibiotics intervene in the functioning of the ribosome of a bacterium , a large and complex biochemical machine that carries out protein biosynthesis by translating messenger RNA into proteins. Macrolide antibiotics block the exit tunnel of the ribosome, preventing the synthesis of essential bacterial proteins. In 2007 the Pande Laboratory received a grant to study and develop new antibiotics. In 2008 they used Folding @ home to investigate the inside of this tunnel and how certain molecules can influence it. The full structure of the ribosome was not determined until 2011, and Folding @ home has also simulated ribosomal proteins as many of their functions are largely unknown.

Computing power

Computing power development
date Computing power
09/16/2007 1 PetaFLOPS
05/07/2008 2 PetaFLOPS
08/20/2008 3 PetaFLOPS
09/28/2008 4 PetaFLOPS
07/19/2016 100 PetaFLOPS
March 20, 2020 470 PetaFLOPS
April 13, 2020 2.4 ExaFLOPS
April 16, 2020 2.6 ExaFLOPS
05/21/2020 2.2 ExaFLOPS
07/03/2020 1.5 ExaFLOPS
07/25/2020 1.3 ExaFLOPS
08/02/2020 1.2 ExaFLOPS

2007 to 2016

Between June 2007 and June 2011, the computing power of all computers involved in the Folding @ home project exceeded the performance of the fastest supercomputer, the TOP500 . However, it was dwarfed by the K computer in November 2011 and the Blue Gene / Q computer in June 2012 . On September 16, 2007, thanks largely to the participation of PlayStation 3 consoles, the Folding @ home project officially reached a level of performance that was higher than a native PetaFLOPS , making it the first computer system ever to achieve this. On the same day, the record was entered in the Guinness Book of Records .

On May 7, 2008, the project reached a sustained level of performance that was higher than two native PetaFLOPS, followed by the three and four PetaFLOPS milestones on August 20, 2008 and September 28, 2008, respectively. On July 19, 2016 it was announced that that the computing power of 100 x86 - PetaFLOPS has been exceeded.

COVID-19 pandemic (2020)

Before the outbreak of the COVID-19 pandemic , around 30,000 users took part in the project. In a Financial Times video on YouTube on April 7, 2020, Gregory Bowman said that over 700,000 new users had been gained as part of the pandemic. Among other things , Nvidia called on computer gamers to contribute their GPU computing power.

On March 20, 2020, Folding @ Home announced that it had computing power of more than 470 x86 - PetaFLOPS , which clearly exceeded the fastest supercomputer to date - the IBM Summit with 148 PetaFLOPS . On April 13th, the project had a computing power of over 2.4 x86 ExaFLOPS and over 1.4 million users, making it faster than all TOP500 supercomputers in the world put together or 15 times faster than IBM Summit. On April 16, the computing power passed the mark of 2.6 x86 -ExaFLOPS. On May 21, 2020, Folding @ home still had a computing power of 2.2 x86 - ExaFLOPS . In the meantime, it has been further reduced from 1.5 x86 -ExaFLOPS on July 3, 2020 to 1.3 x86 -ExaFLOPS on July 25, 2020. On August 2, 2020, the computing power was 1.2 ExaFLOPS.

software

Anyone using a PC with Windows , macOS or Linux can download a client program that works as a service in the background. The current version (7.6.3) supports single-core and multi-core processors as well as graphics cards from Nvidia and AMD . The first client in 2000 was a screen saver that ran while the computer was not otherwise in use.

Professional software developers are responsible for most of the code in the Folding @ home software, on both the client and server sides. The development team includes programmers from Nvidia, ATI, Sony and Cauldron Development.

Clients can only be downloaded from the official Folding @ home website or their commercial partners and only interact with Folding @ home computer files. They exchange data exclusively with the Folding @ home web servers (via port 8080 , alternatively 80). Communication is verified using 2048-bit digital signatures.

The client, GROMACS , various cores and the graphical user interface (GUI) of the client are open source .

Work unit / work unit

A unit of work is the protein data that the client has to process. Units of work are a fraction of the simulation between states in a Markov state model . After the unit of work is downloaded and fully processed by the computer, it is returned to the Folding @ home servers, which then award credits to the user. This cycle repeats itself automatically. All work items have associated deadlines, and if these deadlines are exceeded, those work items are automatically redistributed to another user. Since protein folding is done serially and many units of work are generated from their respective predecessors, the simulation process can run normally even if a unit of work has not been sent back to the server after a reasonable period of time.

Before being released to the public, the units of work go through several quality assurance steps to prevent problematic units from becoming fully available. These test phases include internal, beta and advanced phases before a final full release via Folding @ home takes place.

Folding @ home's units of work are normally processed only once, except in the rare event that processing errors occur. If this case occurs with three different users, the unit is automatically removed from the distribution.

Cores

Special molecular dynamics programs, which are called FahCores and are often abbreviated as Core (in German Kern ), perform the calculations on the work unit as a background process . Cores are scientific computer programs that are specifically designed to perform calculations from the modification and optimization of molecular dynamics simulations. A large majority of the cores used by Folding @ home are based on GROMACS . Less actively used cores are ProtoMol and SHARPEN . Folding @ home also used AMBER , CPMD , Desmond and TINKER , but these have since been discontinued. Some of these cores can represent the water models known in computational chemistry , in which the surrounding solvent (usually water) is modeled atom by atom ( Explicit Solvation method). Other cores perform implicit solvation methods in which the solvent is treated as a mathematical continuum. The core is separate from the client so that the scientific methods can be updated automatically without the need for a client update. The cores regularly create calculation checkpoints so that if the calculation is interrupted, it can be continued at that point.

GPU support

Folding @ home viewer that visualizes client activities

Depending on the setting, the Folding @ home client can use the CPU as well as the GPU for the calculation . Graphics cards from Nvidia and AMD are supported. CUDA technology is required for Nvidia graphics cards (from G80 with GeForce driver from 174.55). AMD graphics cards are supported from the HD5000 series. The graphics units of all APUs from AMD, whether desktop or notebook, can now also be used. The V7 client uses the OpenCL standard .

PlayStation 3

From March 2007 to November 2012, Folding @ home used the computing power of the PlayStation 3 . At the time of its introduction, the main yielded Cell - stream processing a twenty-fold speed increase over PCs for some calculations. The high speed and efficiency of the PS3 opened up further possibilities for optimization in accordance with Amdahl's law (a mathematical equation with which the total acceleration of a program when executed in parallel based on the parallel proportion of a program and the number of processors can be calculated). They significantly changed the relationship between computational efficiency and overall accuracy, allowing more complex molecular models to be used with little additional computational cost. This enabled Folding @ home to carry out biomedical calculations that otherwise would not have been mathematically possible.

The PS3 client was developed in collaboration with Sony and Pande Lab and was first published as a standalone client on March 23, 2007. Upon release, Folding @ home became the first distributed computer project to use the PS3. On 18 September of the following year, the PS3 client was when it was introduced to a passage of Life with PlayStation (a former Multimedia - application software of the PlayStation Network ). Unlike clients that run on PCs, users could not perform other activities on their PS3 while Folding @ home was running. The uniform console environment of the PS3 made technical support easier and made Folding @ home more user-friendly. The PS3 also had the ability to quickly stream data to its graphics processor , which was used for real-time visualization of current protein dynamics at the atomic level.

On November 6, 2012, Sony ended support for Folding @ home for the PS3 client and other services available at Life with PlayStation . During its five-year and seven-month term, more than 15 million users provided more than 100 million hours of computing time for Folding @ home and provided significant support to the project in research into diseases. After talking to Pande Lab, Sony decided to end the application. Vijay Pande saw the PlayStation 3 client as an important development step for the project.

Client V7 for Mac, Windows & Linux

The V7 client is the seventh and latest generation of the Folding @ home client software and represents a complete new version and standardization of the earlier clients for the Windows, MacOS and Linux operating systems.

As of March 2020, the minimum system requirement for the Folding @ home client is a Pentium 4 1.4 GHz CPU .

An example image of the V7 client in novice mode under Windows 7.

It was released on March 22, 2012. Like its predecessors, V7 Folding @ home can run in the background with a very low priority so that other applications can use the CPU resources as required. It's designed to be more user-friendly for beginners to install, start up, and operate, and provide researchers with greater scientific flexibility than previous clients. V7 uses Trac to manage its bug tickets so that users can see the development process and provide feedback.

V7 consists of four integrated elements. The user typically interacts with the open source GUI of V7, called FAHControl. This has the user interface modes "Novice", "Advanced" and "Expert" and offers the possibility to monitor, configure and control many remote folding clients from one computer.

FAHControl controls the FAHClient, a back-end application that in turn manages each FAH slot (or slot). Each slot acts as a replacement for the previously different Folding @ home v6 uniprocessor, SMP or GPU computer clients, as it can download, process and upload work units independently of one another. The FAHViewer function, which is modeled on the viewer of the PS3, shows, if available, a real-time 3D representation of the protein currently being processed.

Chrome

In 2014, a client for the Google Chrome and Chromium web browsers was released that allows users to run Folding @ home in their web browser. The client used Google's native client functionality (NaCl) on Chromium-based web browsers to execute the Folding @ Home code at near native speed in a sandbox on the user's computer. Due to the leakage of NaCl and changes to Folding @ Home, the web client was finally switched off in June 2019.

Android

In July 2015, a client for Android phones was released on Google Play for devices with Android 4.4 KitKat or newer.

On February 16, 2018, the Android client, which was offered in cooperation with Sony , was removed from the Google Play Store. Plans have been announced to offer an open source alternative in the future, but as of August 2020 there was no new Android client.

Motivational incentives

As with many other projects that use distributed computing, Folding @ home also creates statistics in the form of a point system on the computing power contributed. On average, between 500 and 550 projects run under Folding @ home, with individual base points for each project that are defined by a reference PC. For each completed work unit of a project, the user receives the base points provided for it. However, these points can be multiplied (bonus points) the faster a work unit is completed. The prerequisite for receiving points is a registration with the project with a user name. In order to receive bonus points, a so-called "passkey" is also required.

Each user can choose whether his computing power is counted anonymously (no points), only under his user name or also for a team.

If a user does not form a new team or does not join an existing team, that user automatically becomes part of the “standard” team. This "standard" team has the team number "0". The statistics are collected for this “standard” team as well as for specially named teams.

Results

A total of 224 scientific publications (as of May 15, 2020) were published as a direct result of Folding @ home.

Related projects

Rosetta @ home , Predictor @ home and POEM @ home were or are projects that have the same goal but use different methods. Foldit, for example, is an experimental computer game that aims to help scientists optimize proteins. The DreamLab project enables smartphones to provide processor capacities for medical research at Imperial College London while they are charging .

Rosetta @ home

Rosetta @ home is a distributed computing project for predicting protein structures and is one of the most accurate tertiary structure predictors. The conformational states from Rosetta software can be used to initialize a Markov state model as the starting point for Folding @ home simulations. Conversely, the algorithms for structure prediction from thermodynamic and kinetic models as well as the sampling aspects of protein folding simulations can be improved. Since Rosetta only tries to predict the final state of folding and not how the folding will take place, Rosetta @ home and Folding @ home are complementary and address very different molecular issues.

Foldit

The aim of Foldit is to obtain a protein that is as “ folded ” as possible, ie. H. a model of the protein in the state of energy minimum. That is the form in which it occurs in nature. However, no previous knowledge is required, the evaluation is done by the program. The protein manipulation options available to the player are explained in a series of tutorial puzzles. For the game, a graphic equivalent of the protein structure is displayed, which the player can change using various tools. If the structure is changed, the program calculates a score based on how well the protein is folded. For each puzzle a high score is calculated for both individual and group solutions, which changes in real time. Foldit is an attempt to apply the natural human 3-D pattern recognition skills to this problem. Current puzzles are based on well-understood proteins. By examining how players intuitively approach these puzzles, the scientists are trying to improve existing protein folding software.

Anton

Anton is a supercomputer that for molecular dynamics - simulations was built. In October 2011, Anton and Folding @ home were the two most powerful molecular dynamics systems. Anton is unique in his ability to generate individual ultra-long, computationally intensive molecular trajectories such as one in 2010 that reached the millisecond range. These long trajectories can be especially helpful for some types of biochemical problems. However, Anton does not use Markov State Models (MSM) for the analysis. In 2011, the Pande laboratory constructed an MSM from two 100 microsecond Anton simulations and found alternative folding pathways that were not visible by Anton's traditional analysis. They concluded that there was little difference between MSMs constructed from a limited number of long lanes and those constructed from many shorter lanes. In June 2011, Folding @ home added sampling to an Anton simulation to better determine how their methods compare to Anton's methods. In contrast to the shorter tracks from Folding @ home, which are better suited for distributed calculations and other parallelization methods, longer tracks do not require adaptive sampling in order to adequately sample the phase space of the protein. For this reason it is possible that a combination of the simulation methods from Anton and Folding @ home would allow a more thorough scan of this room.

Web links

Commons : Folding @ home  - collection of pictures, videos and audio files

Individual evidence

  1. ^ Vijay S. Pande, Kyle Beauchamp, Gregory R. Bowman: Everything you wanted to know about Markov State Models but were afraid to ask . In: Methods (San Diego, Calif.) . tape 52 , no. 1 , September 2010, ISSN  1046-2023 , p. 99-105 , doi : 10.1016 / j.ymeth.2010.06.002 , PMID 20570730 , PMC 2933958 (free full text).
  2. ^ Maria Temming: You can help fight the coronavirus. All you need is a computer. In: Science News. March 25, 2020, accessed March 26, 2020 (American English).
  3. a b Folding @ Home with 2.4 exaflops faster than the top 500 supercomputers. April 14, 2020, accessed April 14, 2020 .
  4. New Folding @ home software with the option to prioritize COVID-19 projects
  5. ^ Vijay Pande: History of Folding @ Home. Stanford University, April 14, 2007, accessed March 23, 2020 .
  6. Public Launch. Stanford University, September 19, 2000, accessed on March 22, 2020 (English): "We released our software to the public and very soon after we had thousands of computers donating otherwise unused computer power."
  7. ^ Julia Evangelou Strait: Bowman leading international supercomputing project. Washington University School of Medicine in St. Louis, February 26, 2019, accessed on March 19, 2020 : “ With this networked computing power, Folding @ home is, essentially, one of the world's largest supercomputers. "
  8. Folding @ Home FAQ. Retrieved March 6, 2020 .
  9. ^ Vijay Pande: Folding @ home and Simbios. April 23, 2008, accessed on March 23, 2020 : "Second, we're also starting to make large data sets from Folding @ home available to others. You can see some of the first data sets on this project page, and we expect to put more data up as time goes on. Folding @ home donors have generated wonderful data sets that aren't possible to generate by other means, and our hope is to publish them so that other scientists can data mine them for other purposes as well. "
  10. ^ Vijay Pande: Changes to F @ h Website. October 25, 2011, accessed on March 23, 2020 (English): "We have been making these available in general on request and in cases where people ask for data sets repeatedly (eg https://simtk.org/home/foldvillin ) , we make them available on a website linked from folding.stanford.edu. "
  11. ^ Matthew P. Harrigan, Mohammad M. Sultan, Carlos X. Hernández, Brooke E. Husic, Peter Eastman: MSMBuilder: Statistical Models for Biomolecular Dynamics . In: Biophysical Journal . tape 112 , no. 1 , January 10, 2017, ISSN  1542-0086 , p. 10–15 , doi : 10.1016 / j.bpj.2016.10.042 , PMID 28076801 , PMC 5232355 (free full text).
  12. SimTK: MSMBuilder: News. Retrieved March 18, 2020 .
  13. ^ Biophysical Society Names Five 2012 Award Recipients. (No longer available online.) August 29, 2011, archived from the original on September 21, 2012 ; accessed on March 18, 2020 (English).
  14. Award. (No longer available online.) Archived from the original on September 21, 2012 ; accessed on March 18, 2020 .
  15. ^ C. Pratt, K. Cornely: Thermodynamics. Essential biochemistry . Wiley-VCH, 2004, ISBN 978-0-471-39387-0 .
  16. Vijay Pande: When and where do Proteins fold? Stanford University, August 16, 2007, accessed March 23, 2020 .
  17. G. Zhang, Z. Ignatova: Folding at the birth of the nascent chain: coordinating translation with co-translationally folding. In: Current opinion in structural biology. Volume 21, Number 1, February 2011, pp. 25-31, doi: 10.1016 / j.sbi.2010.10.008 , PMID 21111607 (review).
  18. B. van den Berg, R. Wain, CM Dobson, RJ Ellis: Macromolecular crowding perturbs protein refolding kinetics: implications for folding inside the cell. In: The EMBO Journal . Volume 19, number 15, August 2000, pp. 3870-3875, doi: 10.1093 / emboj / 19.15.3870 , PMID 10921869 , PMC 306593 (free full text).
  19. ^ S. Al-Karadaghi: Torsion Angles and the Ramachandran Plot. In: proteinstructures.com. Retrieved March 17, 2020 .
  20. F. Marinelli, F. Pietrucci, A. Laio, S. Piana: A kinetic model of trp-cage folding from multiple biased molecular dynamics simulations. In: PLoS Computational Biology. Volume 5, number 8, August 2009, p. E1000452, doi: 10.1371 / journal.pcbi.1000452 , PMID 19662155 , PMC 2711228 (free full text).
  21. : So much more to know. In: Science . Volume 309, Number 5731, July 2005, pp. 78-102, doi: 10.1126 / science.309.5731.78b , PMID 15994524 .
  22. a b H. Ecroyd, YES Carver: Unraveling the mysteries of protein folding and misfolding. In: IUBMB life. Volume 60, Number 12, December 2008, pp. 769-774, doi: 10.1002 / iub.117 , PMID 18767168 (review).
  23. ^ A b Y. Chen, F. Ding, H. Nie, AW Serohijos, S. Sharma, KC Wilcox, S. Yin, NV Dokholyan: Protein folding: then and now. In: Archives of biochemistry and biophysics. Volume 469, number 1, January 2008, pp. 4-19, doi: 10.1016 / j.abb.2007.05.014 , PMID 17585870 , PMC 2173875 (free full text).
  24. LM Luheshi, DC Crowther, CM Dobson: Protein misfolding and disease: from the test tube to the organism. In: Current opinion in chemical biology. Volume 12, Number 1, February 2008, pp. 25-31, doi: 10.1016 / j.cbpa.2008.02.011 , PMID 18295611 (review).
  25. CD Snow, EJ Sorin, YM Rhee, VS Pande: How well can simulation predict protein folding kinetics and thermodynamics? In: Annual review of biophysics and biomolecular structure. Volume 34, 2005, pp. 43-69, doi: 10.1146 / annurev.biophys.34.040204.144447 , PMID 15869383 (review).
  26. ^ A b V. A. Voelz, GR Bowman, K. Beauchamp, VS Pande: Molecular simulation of ab initio protein folding for a millisecond folder NTL9 (1-39). In: Journal of the American Chemical Society . Volume 132, number 5, February 2010, pp. 1526-1528, doi: 10.1021 / ja9090353 , PMID 20070076 , PMC 2835335 (free full text).
  27. A. Verma, S .M. Gopal, A. Schug, JS Oh, KV Klenin, KH Lee, W. Wenzel: Massively Parallel All Atom Protein Folding in a Single Day In: Advances in Parallel Computing. Volume 15, 2008, pp. 527-534, ISBN 978-1-58603-796-3 .
  28. VS Pande, I. Baker, J. Chapman, SP Elmer, S. Khaliq, SM Larson, YM Rhee, MR Shirts, CD Snow, EJ Sorin, B. Zagrovic: Atomistic protein folding simulations on the submillisecond time scale using worldwide distributed computing. In: Biopolymers. Volume 68, Number 1, January 2003, pp. 91-109, doi: 10.1002 / bip.10219 , PMID 12579582 .
  29. ^ A b G. R. Bowman, VA Voelz, VS Pande: Taming the complexity of protein folding. In: Current opinion in structural biology. Volume 21, number 1, February 2011, pp. 4-11, doi: 10.1016 / j.sbi.2010.10.006 , PMID 21081274 , PMC 3042729 (free full text) (review).
  30. John D. Chodera, William C. Swope, Jed W. Pitera, Ken A. Dill: Long ‐ Time Protein Folding Dynamics from Short ‐ Time Molecular Dynamics Simulations. In: Multiscale Modeling & Simulation. Volume 5, 2006, p. 1214, doi: 10.1137 / 06065146X .
  31. RB Best: Atomistic molecular simulations of protein folding. In: Current opinion in structural biology. Volume 22, number 1, February 2012, pp. 52-61, doi: 10.1016 / j.sbi.2011.12.001 , PMID 22257762 (review).
  32. ^ TJ Lane, Gregory Bowman, Robert McGibbon, Christian Schwantes, Vijay Pande, Bruce Borden: Folding @ home Simulation FAQ. Folding @ home, September 21, 2012, accessed March 17, 2020 .
  33. ^ A b V. S. Pande, K. Beauchamp, GR Bowman: Everything you wanted to know about Markov State Models but were afraid to ask. In: Methods. Volume 52, number 1, September 2010, pp. 99-105, doi: 10.1016 / j.ymeth.2010.06.002 , PMID 20570730 , PMC 2933958 (free full text) (review).
  34. ^ GR Bowman, DL Ensign, VS Pande: Enhanced modeling via network theory: Adaptive sampling of Markov state models. In: Journal of chemical theory and computation. Volume 6, number 3, 2010, pp. 787-794, doi: 10.1021 / ct900620b , PMID 23626502 , PMC 3637129 (free full text).
  35. Vijay Pande: FAHcon 2012: Thinking about how far FAH has come. In: typepad.com. Folding @ home, June 8, 2012, accessed March 17, 2020 .
  36. ^ Brooke E. Husic, Vijay S. Pande: Markov State Models: From an Art to a Science . In: Journal of the American Chemical Society . tape 140 , no. 7 , February 21, 2018, ISSN  1520-5126 , p. 2386-2396 , doi : 10.1021 / jacs.7b12191 , PMID 29323881 .
  37. ^ CD Snow, H. Nguyen, VS Pande, M. Gruebele: Absolute comparison of simulated and experimental protein-folding dynamics. In: Nature . Volume 420, number 6911, November 2002, pp. 102-106, doi: 10.1038 / nature01160 , PMID 12422224 .
  38. KA Beauchamp, DL Ensign, R. Das, VS Pande: Quantitative comparison of villin headpiece subdomain simulations and triplet-triplet energy transfer experiments. In: Proceedings of the National Academy of Sciences . Volume 108, number 31, August 2011, pp. 12734-12739, doi: 10.1073 / pnas.1010880108 , PMID 21768345 , PMC 3150881 (free full text).
  39. Simulation of millisecond protein folding: NTL9 (from Folding @ home). Simulating protein folding on the millisecond timescale has been a major challenge for many years. In a recent paper ( http://pubs.acs.org/doi/abs/10.1021/j ...), Folding @ home researchers Vincent Voelz, Greg Bowman, Kyle Beauchamp, and Vijay Pande have broken this barrier. This is a movie of one of the trajectories that folded (ie started unfolded and ended up in the folded state). Stanford University, January 18, 2010, accessed March 23, 2020 .
  40. TH Click, D. Ganguly, J. Chen: Intrinsically disordered proteins in a physics-based world. In: International Journal of Molecular Sciences . Volume 11, number 12, 2010, pp. 5292-5309, doi: 10.3390 / ijms11125292 , PMID 21614208 , PMC 3100817 (free full text) (review).
  41. Which diseases or biomedical problems are you currently studying? Stanford University, September 12, 2012, accessed March 18, 2020 .
  42. ^ Greg Bowman: Folding Forum • Login. August 22, 2011, accessed on March 18, 2020 (English): “This A3 project for windows and Linux clients aims to characterize the dynamics of RNase H, a key component of HIV. By understanding the role of dynamics in its mechanism, we hope to be better able to design drugs to deactivate this enzyme. "
  43. Hana Robson Marsden, Itsuro Tomatsu, Alexander Kros: Model systems for membrane fusion . In: Chemical Society Reviews . tape 40 , no. 3 , February 22, 2011, ISSN  1460-4744 , p. 1572-1585 , doi : 10.1039 / C0CS00115E .
  44. ^ Gregory R Bowman, Xuhui Huang, Vijay S Pande: Network models for molecular kinetics and their initial applications to human health . In: Cell research . tape 20 , no. June 6 , 2010, ISSN  1001-0602 , p. 622–630 , doi : 10.1038 / cr.2010.57 , PMID 20421891 , PMC 4441225 (free full text).
  45. ^ Peter M. Kasson, Afra Zomorodian, Sanghyun Park, Nina Singhal, Leonidas J. Guibas: Persistent voids: a new structural metric for membrane fusion . In: Bioinformatics . tape 23 , no. 14 , July 15, 2007, ISSN  1367-4803 , p. 1753–1759 , doi : 10.1093 / bioinformatics / btm250 ( oup.com [accessed March 18, 2020]).
  46. Peter M. Kasson, Daniel L. Ensign, Vijay S. Pande: Combining molecular dynamics with Bayesian analysis to predict and evaluate ligand-binding mutations in influenza hemagglutinin. In: Journal of the American Chemical Society . tape 131 , no. 32 , August 19, 2009, ISSN  0002-7863 , p. 11338–11340 , doi : 10.1021 / ja904557w , PMID 19637916 , PMC 2737089 (free full text).
  47. ^ Protein folding and viral infection. Retrieved March 18, 2020 .
  48. Christian Gruber, Georg Steinkellner: Wuhan coronavirus 2019-nCoV - what we can find out on a structural bioinformatics level . In: Innophore . January 29, 2020. doi : 10.6084 / m9.figshare.11752749 .
  49. Sean Fleming: These researchers want to use your computer in the search for a COVID-19 vaccine. World Economic Forum, March 17, 2020, accessed on March 22, 2020 .
  50. ^ Kiona Smith: SETI @ Home Is Over; The Fight Against COVID-19 Coronavirus Is Just Beginning. Forbes Magazine, March 15, 2020, accessed on March 22, 2020 (English): "Folding @ Home (a program similar to SETI @ Home that focuses on disease research - specifically how proteins fold) just rolled out" an initial wave of projects "That simulate how proteins from SARS-CoV-2 (the virus that causes COVID-19) work and how they interact with human cells."
  51. Greg Bowman: FOLDING @ HOME TAKES UP THE FIGHT AGAINST COVID-19/2019-NCOV. February 27, 2020, accessed on March 18, 2020 .
  52. Folding @ home Turns Its Massive Crowdsourced Computer Network Against COVID-19. March 16, 2020, accessed March 18, 2020 (American English).
  53. John Chodera: FOLDING @ HOME UPDATE ON SARS-COV-2 (10 MAR 2020). March 10, 2020, accessed on March 18, 2020 .
  54. Julia Evangelou Strait: Crowdsourced supercomputing project sets sights on coronavirus. Washington University School of Medicine in St. Louis, March 19, 2020, accessed March 19, 2020 .
  55. Greg Bowman: CORONA VIRUS - WHAT WE'RE DOING AND HOW YOU CAN HELP IN SIMPLE TERMS. Washington University, March 15, 2020, accessed on March 22, 2020 (English): "We're simulating the dynamics of COVID-19 proteins to hunt for new therapeutic opportunities."
  56. ^ G Brent Irvine; Omar M El-Agnaf; Ganesh M Shankar; Dominic M Walsh: Protein Aggregation in the Brain: The Molecular Basis for Alzheimer's and Parkinson's Diseases . Molecular Medicine (review). 14 (7-8): 451-464, 2008.
  57. Claudio Soto; Lisbell D. Estrada: Protein Misfolding and Neurodegeneration . Archives of Neurology (review). 65 (2): 184-189, 2008.
  58. Robin Roychaudhuri; Mingfeng Yang; Minako M. Hoshi; David B. Teplow: Amyloid β-Protein Assembly and Alzheimer's Disease . Journal of Biological Chemistry. 284 (8): 4749-53, 2008.
  59. a b Nicholas W. Kelley; V. Vishal; Grant A. Krafft; Vijay S. Pande: Simulating oligomerization at experimental concentrations and long timescales: A Markov state model approach . Journal of Chemical Physics. 129 (21): 214707, 2008.
  60. Aabgeena Naeem; Naveed Ahmad Fazili: Defective Protein Folding and Aggregation as the Basis of Neurodegenerative Diseases: The Darker Aspect of Proteins . Cell Biochemistry and Biophysics (review). 61 (2): 237-50, 2011.
  61. Gregory R Bowman; Xuhui Huang; Vijay S Pande: Network models for molecular kinetics and their initial applications to human health . Cell Research (review). 20 (6): 622-630, 2010.
  62. ^ Vijay Pande: New FAH results on possible new Alzheimer's drug presented. Stanford University, December 18, 2008, accessed March 23, 2020 .
  63. J. Rajadas, CW Liu, P. Novick, NW Kelley, M. Inayathullah, MC Lemieux, VS Pande: Rationally designed turn promo ting mutation in the amyloid-β peptide sequence stabilizes oligomers in solution. In: PLOS ONE . Volume 6, number 7, 2011, p. E21776, doi: 10.1371 / journal.pone.0021776 , PMID 21799748 , PMC 3142112 (free full text).
  64. a b c Pande lab: Folding @ home Diseases Studied FAQ . Folding @ home. foldingathome.org. Archived from the original on September 21, 2012. Retrieved March 18, 2020.
  65. Walker FO: Huntington's disease . Lancet. 369 (9557): 218-28, 2007, p. 220
  66. Nicholas W. Kelley; Xuhui Huang; Stephen Tam; Christoph Spiess; Judith Frydman; Vijay S. Pande: The predicted structure of the headpiece of the Huntingtin protein and its implications on Huntingtin aggregation . Journal of Molecular Biology. 388 (5): 919-27, 2009. doi: 10.1016 / j.jmb.2009.01.032.
  67. ^ Susan W Liebman: Protein folding: Sticky N17 speeds huntingtin pile-up. Springer Nature Limited, January 2010, accessed on March 23, 2020 : "Aggregation of huntingtin protein with an expanded polyglutamine region is enhanced by its 17-residue N-terminal domain, which binds to itself and to the polyglutamine region. This enhancement is inhibited when the N-terminal domain binds to the chaperonin TRiC. "
  68. a b c d Pande lab: Folding @ home Diseases Studied FAQ , 2012, accessed on March 18, 2020.
  69. ^ M Hollstein; D Sidransky; B Vogelstein; CC Harris (1991). p53 mutations in human cancers . Science. 253 (5015): 49-53.
  70. Lillian T Chong: Dimerization of the p53 Oligomerization Domain: Identification of a Folding Nucleus by Molecular Dynamics Simulations. ScienceDirect, January 28, 2005, accessed March 23, 2020 .
  71. Lillian T. Chong; William C. Swope; Jed W. Pitera; Vijay S. Pande: Kinetic Computational Alanine Scanning: Application to p53 Oligomerization . Journal of Molecular Biology, 2005. 357 (3): 1039-1049.
  72. Almeida MB, do Nascimento JL, Herculano AM, Crespo-López ME: Molecular chaperones: toward new therapeutic tools . Journal of Molecular Biology, 65 (4): 239-43, 2011.
  73. ^ Stanford University Medical Center: Scientists boost potency, reduce side effects of IL-2 protein used to treat cancer. MedicalXPress, March 18, 2012, accessed on March 23, 2020 .
  74. Aron M. Levin; Darren L. Bates; Aaron M. Ring; Carsten War; Jack T. Lin; Leon Su; Ignacio Moraga; Miro E. Raeber; Gregory R. Bowman; Paul Novick; Vijay S. Pande; C. Garrison Fathman; Onur Boyman; K. Christopher Garcia: Exploiting a natural conformational switch to engineer an interleukin-2 'superkine' . Nature. 484 (7395): 529-33, 2012.
  75. Sanghyun Park, Randall J. Radmer, Teri E. Klein: A new set of molecular mechanics parameters for hydroxyproline and its use in molecular dynamics simulations of collagen ‐ like peptides. September 16, 2005, accessed March 23, 2020 .
  76. ^ Vijay Pande: New methods for computational drug design . Folding @ home. typepad.com. Archived on September 21, 2012. Retrieved March 20, 2020.
  77. Guha Jayachandran; MR Shirts; S. Park; VS Pande: Parallelized-Over-Parts Computation of Absolute Binding Free Energy with Docking and Molecular Dynamics . Journal of Chemical Physics. 125 (8): 084901, 2006.
  78. Chun Song; Shen Lim; Joo Tong: Recent advances in computer-aided drug design . Briefings in Bioinformatics (review). 10 (5): 579-91, 2009.
  79. Pande Laboratory: Project 10721 Description home Folding @. foldingathome.org. Archived on September 21, 2012. Retrieved March 20, 2020.
  80. ^ A b Gregory Bowman: Searching for new drug targets. Folding @ home. foldingathome.org. Archived on September 21, 2012. Retrieved March 20, 2020.
  81. Gregory R. Bowman; Phillip L. Geissler: Equilibrium fluctuations of a single folded protein reveal a multitude of potential cryptic allosteric sites . PNAS. 109 (29): 11681-6, 2012.
  82. Paula M. Petrone; Christopher D. Snow; Del Lucent; Vijay S. Pande: Side-chain recognition and gating in the ribosome exit tunnel . Proceedings of the National Academy of Sciences. 105 (43): 16549-54, 2008.
  83. ^ Prof. Stefan Voelz: Folding @ home Project Description. Retrieved March 23, 2020 (English).
  84. What are native FLOPS? In: FAQs. Folding @ Home, accessed on March 23, 2020 (English): “We refer to the FLOP count on a given hardware as the native FLOP count. For example, an exponential on a GPU is one native GPU FLOP but many native x86 FLOPS. "
  85. Crossing the petaflop barrier. In: Folding @ home A blog all about Folding @ home, from its Director, Prof. Vijay Pande. September 16, 2007, accessed March 6, 2020 .
  86. Most powerful distributed computing network. Guinness World Records Limited, September 16, 2007, accessed on March 22, 2020 : "On September 16, 2007 Folding @ home, a distributed computing network operating from Stanford University (USA) achieved a computing power of 1 petaflop - or 1 quadrillion floating point operations per second. "
  87. Guest Recorder: 3 PetaFLOP barrier - PetScience. LongeCity - Advocacy & Research for Unlimited Lifespans, August 19, 2008, accessed March 6, 2020 .
  88. Folding @ Home: A significant milestone: 100 petaFLOPS. July 19, 2016, accessed March 6, 2020 .
  89. Michael Günsch: Rosetta @ home: Active users doubled due to the fight against coronavirus .
  90. Coronavirus: how your computer could help find a vaccine | FT. Financial Times, April 7, 2020, accessed on April 7, 2020 : "In the past three week, though, we had over 700,000 people join the project."
  91. Connor Sheridan March 13, 2020: Nvidia's calling on gaming PC owners to put their systems to work fighting COVID-19. In: gamesradar.com. gamesradar, accessed on April 26, 2020 (English).
  92. Greg Bowman: Amazing! @foldingathome now has over 470 petaFLOPS of compute power. To put that in perspective, that's more than 2x the peak performance of the Summit super computer! In: @drGregBowman. Twitter, March 20, 2020, accessed March 20, 2020 .
  93. Folding @ home stats report. Retrieved April 16, 2020 .
  94. Folding @ home stats report. Retrieved May 21, 2020 .
  95. Folding @ home stats report. Retrieved July 3, 2020 .
  96. Folding @ home stats report. Retrieved July 25, 2020 .
  97. Folding @ home stats report. Retrieved August 2, 2020 .
  98. Michael Shirts: Screen Savers of the World Unite. 2000, accessed on March 23, 2020 .
  99. How does FAH code development and sysadmin get done? Retrieved March 18, 2020 .
  100. FAQ Folding at Home. Pande Lab, September 21, 2012, accessed March 18, 2020 .
  101. OpenSource - Folding @ home .
  102. WHAT IS A WORK UNIT? Retrieved March 19, 2020 .
  103. ^ Running Folding @ Home. Retrieved on March 23, 2020 (English): "These unfinished work units" expire "and are reassigned to new machines. You will still receive credit for all WUs completed and uploaded prior to the timeout (formerly preferred deadline). "
  104. More transparency in testing. Retrieved March 19, 2020 .
  105. Folding Forum • View topic - Gromacs cannot continue further. Retrieved March 19, 2020 .
  106. a b Adam Beberg, Daniel Ensign, Guha Jayachandran, Siraj Khaliq, Vijay Pande: Folding @ home: Lessons from eight years of volunteer distributed computing. (PDF) In: hicomb.org. 2009 IEEE International Symposium on Parallel & Distributed Processing. Proceedings, 2009, accessed March 18, 2020 .
  107. Pande lab: OPENSOURCE. In: webcitation.org. August 3, 2012, accessed March 18, 2020 .
  108. Pande lab: Folding @ home Frequently Asked Questions (FAQ) Index. In: webcitation.org. September 21, 2012, accessed March 18, 2020 .
  109. ^ Vijay Pande: Update on new FAH cores and clients. In: folding.typepad.com. Folding @ home, September 25, 2009, accessed March 18, 2020 .
  110. MS Friedrichs, P. Eastman, V. Vaidyanathan, M. Houston, S. Legrand, AL Beberg, DL Ensign, CM Bruns, VS Pande: Accelerating molecular dynamic simulation on graphics processing units. In: Journal of computational chemistry. Volume 30, Number 6, April 2009, pp. 864-872, doi: 10.1002 / jcc.21209 , PMID 19191337 , PMC 2724265 (free full text).
  111. ^ Pande lab: Folding @ home Petaflop Initiative (FPI). In: webcitation.org. August 19, 2012, accessed March 18, 2020 .
  112. ComputerBase: Folding @ Home Client for Nvidia GPUs is out
  113. Futures in Biotech 27: Folding @ home at 1.3 Petaflops. In: Castroller.com. CastRoller, December 28, 2007, accessed March 18, 2020 .
  114. ^ A b E. Luttmann, DL Ensign, V. Vaidyanathan, M. Houston, N. Rimon, J. Øland, G. Jayachandran, M. Friedrichs, VS Pande: Accelerating molecular dynamic simulation on the cell processor and Playstation 3. In : Journal of computational chemistry. Volume 30, Number 2, January 2009, pp. 268-274, doi: 10.1002 / jcc.21054 , PMID 18615421 .
  115. ^ A b David E. Williams: PlayStation's serious side: Fighting disease. In: CNN. October 20, 2006, accessed March 18, 2020 .
  116. Jerry Liao: The Home Cure: PlayStation 3 to Help Study Causes of Cance. In: mb.com. Manila Bulletin Publishing Corporation, March 23, 2007, accessed March 18, 2020 .
  117. Lou Kesten: Week in video-game news: 'God of War II' storms the PS2; a PS3 research project. In: Post-Gazette.com. Pittsburgh Post-Gazette, March 26, 2007, accessed March 18, 2020 .
  118. Elaine Chow: PS3 News Service, Life With Playstation, Now Up For Download. In: Gizmodo.com. Gizmodo, September 18, 2008, accessed March 18, 2020 .
  119. ^ Vijay Pande: Life with Playstation - a new update to the FAH / PS3 client. In: typepad.com. Folding @ home, September 18, 2008, accessed March 18, 2020 .
  120. Eric Lempel: PS3 System Software Update (v4.30). In: PlayStation blog. Sony, October 21, 2012, accessed March 18, 2020 .
  121. ^ Termination of Life with PlayStation. In: Life with PlayStation. Sony, November 6, 2012, accessed March 18, 2020 .
  122. Windows V7 client. Stanford University, August 18, 2012, accessed on March 6, 2020 (English): "We are pleased to say that everything is new in this version, using completely new software coding from the ground up."
  123. Client version 7 now in open beta. March 9, 2011, accessed on March 6, 2020 (English): "I am happy to announce that after many months of development and testing the new version 7 Folding @ home client software is now available for open-beta testing. The V7 client is a complete rewrite of the previous client for Windows, OS-X and Linux with the following goals "
  124. Requirements. Retrieved on March 23, 2020 : "CPU Slot Requirements Windows XP SP3 or newer, 32 or 64 bit Intel P4 1.4 GHz processor or newer, or AMD equivalent (modern multi-core processors recommended)"
  125. Link to GPL in FAHControl. October 3, 2011, accessed March 6, 2020 .
  126. Folding @ home 3D Viewer. Github, accessed on March 23, 2020 (English): "The Folding @ home viewer allows you to visualize protein folding simulations and monitor the status of the simulation work units as they run on your computer. Installing and running the viewer is not necessary to run Folding @ home. "
  127. Folding @ home Chrome Client. Accessed on March 6, 2020 (English): "Our decision to retire the NaCl client was due to a combination of Google deprecating NaCl and infrastructure upgrades at Folding @ home which would have required extra effort to continue to support the NaCl folding client"
  128. Anto Thynell: android-client-overhaul. February 2, 2018, accessed on March 6, 2020 (English): "From the 16th of February 2018, it will no longer be possible to use the Folding @ Home Android client from Sony"
  129. Alternative downloads / clients. Retrieved March 19, 2020 .
  130. Folding @ home: FAQ-Points ( Memento from July 19, 2013 in the Internet Archive )
  131. Vickie Curtis: Patterns of Participation and Motivation in Folding @ home: The Contribution of Hardware Enthusiasts and Overclockers . In: Citizen Science: Theory and Practice . tape 3 , no. 1 , April 27, 2018, ISSN  2057-4991 , p. 5 , doi : 10.5334 / cstp.109 ( citizenscienceassociation.org [accessed March 20, 2020]).
  132. ^ Folding @ Home: Projects Summary
  133. Folding @ home: FAQ Passkey ( Memento from July 16, 2013 in the Internet Archive )
  134. Papers & Results. On FoldingAtHome, accessed May 15, 2020.
  135. ^ Research from Stanford University. Stanford University, accessed March 23, 2020 .
  136. ^ The Science Behind Foldit. In: Foldit. Vanderbilt University, accessed March 19, 2020 .
  137. Jump up Marc F. Lensink, Raúl Méndez, Shoshana J. Wodak: Docking and scoring protein complexes: CAPRI 3rd Edition . In: Proteins: Structure, Function, and Bioinformatics . tape 69 , no. 4 , 2007, ISSN  1097-0134 , p. 704-718 , doi : 10.1002 / prot.21804 ( wiley.com [accessed March 19, 2020]).
  138. What is Rosetta at Home? Retrieved March 19, 2020 .
  139. Vijay Grande: Folding @ home vs. Rosetta @ home. March 5, 2006, accessed on March 19, 2020 : “However, Rosetta and Folding @ Home are addressing very different problems. Rosetta only predicts the final folded state, not how do proteins fold (and Rosetta has nothing to do with protein misfolding). Thus, those methods are not useful for the questions we're interested in and the diseases we're tackling (Alzheimer's Disease and other aggregation related diseases). "
  140. ^ Vijay Pande: Comparison between FAH and Anton's approaches . Folding @ home. typepad.com. Original archived on September 21, 2012. Retrieved March 18, 2020
  141. a b Thomas J. Lane; Gregory R. Bowman; Kyle A Beauchamp; Vincent Alvin Voelz; Vijay S. Pande: Markov State Model Reveals Folding and Functional Dynamics in Ultra-Long MD Trajectories . Journal of the American Chemical Society. 133 (45): 18413-9, 2011.
  142. David E. Shaw; et al .: Millisecond-scale molecular dynamics simulations on Anton. Proceedings of the Conference on High Performance Computing Networking, Storage and Analysis. Pp. 1-11, 2009.
  143. David E. Shaw; et al .: Atomic-Level Characterization of the Structural Dynamics of Proteins . Science. 330 (6002): 341-346, 2010.
  144. Ron O. Dror; Robert M. Dirks; JP Grossman; Huafeng Xu; David E. Shaw: Biomolecular Simulation: A Computational Microscope for Molecular Biology . Annual Review of Biophysics. 41: 429-52, 2012.
  145. David E. Shaw; Martin M. Deneroff; Ron O. Dror; Jeffrey S. Kuskin; Richard H. Larson; John K. Salmon; Cliff Young; Brannon Batson; Kevin J. Bowers; Jack C. Chao; Michael P. Eastwood; Joseph Gagliardo; JP Grossman; C. Richard Ho; Douglas J. Ierardi; et al .: Anton, A Special-Purpose Machine for Molecular Dynamics Simulation . Communications of the ACM. 51 (7): 91-97, 2008.