Tanpaku
"TANPAKU" is the project that is aiming at attacking "protein structure prediction" problem by using "distributed computing" method. This project is developed in collaboration with Yamato lab.(in the Department of Biological Science and Technology) and Takeda lab.(in the Department of Information Sciences) groups at Tokyo University of Science. Now, the project finished a preliminary test within our university and it is open for public volunteers.
What is protein structure prediction?
Proteins play the leading roles in our life
Proteins are essential parts of our life. For example, amylase digests starch to generate energy, hemoglobin transports oxygen around the body and insulin controls the amount of sugar absorbed from the blood.
Protein, amino acid and DNA
The basic units of protein are 20 kinds of amino acids. Proteins are long non-branched polymers composed of linked hundreds of amino acids, where the peptide bonds are formed between the carboxyl group of one amino acid and amino group of another amino acid with the release of one water molecule. The order of its amino acids or sequence determines the properties of the protein.
On the other hand, the DNA base sequence of a gene provides the amino acid sequence information of the protein it encodes. Therefore, gene is a blueprint to make a protein and genome, the complete set of genes in a living organism, is a blueprint of our body as a whole. Because genomic information provides us a lot of clues for solving secrets of life, hundreds of genome sequencing projects of many species are performed around the world. Today, the human genome has been also deciphered!
The relationship between protein structure and function
Given the genomic sequence information, do we understand the whole life? Unfortunately, although we know that a gene corresponds to one of the proteins, it is still difficult to deduce its function from the amino acid sequence.
Protein folds into its unique three-dimensional structure corresponding to its amino acid sequence and then works in our body. In other words, there is a deep relationship between structure of a protein and its function. Therefore, if we can predict a protein structure from its amino acid sequence, this structural information provides very useful information for deducing its protein function.
However, we don't have the method that predicts the folded structure of a protein from its amino acid sequence alone.
Protein structure prediction by computer simulation
There has been an interest in predicting protein structures by computer simulation. Although computational power has greatly increased now, it is still difficult to sample enough conformations directly in a reasonable time to find the folded structure with even state-of-the-art computers.
Under this situation, researchers have proposed various methods to overcome this difficulty. In our laboratory, "Brownian Dynamics (BD) method" that is based on the simulation of Brownian motion has been developed, giving some successful results up to now. This BD method enables us to simulate long time with less computational time compared with conventional simulation methods.
However, toward the final goal of predicting a protein structure from its amino acid sequence, more computational power is required. In order to achieve this requirement, we are improving computational algorithm and also we have paid attention to simulating BD on a more powerful computational resource.
So, we launch the "TANPAKU" project on the distributed computing.
What is distributed computing?
Distributed computing is a computing system which solves a large or difficult problem that requires a huge computational power by giving small parts of the problem to many computers over a network. The most famous distributed computing project is “SETI@home” (SETI stands for Search for Extra-Terrestrial Intelligence) launched at the University of California, Berkeley in May 1999, in which data incoming from the radio telescope is analyzed to search for possible evidence of radio transmissions from extraterrestrial intelligence.
BOINC
"TANPAKU" uses "BOINC" (Berkeley Open Infrastructure for Network Computing) which is a software platform for distributed computing using volunteered computer resources. BOINC helps you participate in more than one distributed computing project, and makes it possible to control the fraction of your computing power dedicated to each project. Today many distributed computing projects including SETI@home are conducted on BOINC platform.
Applications available are:
- Linux/x86
- Windows 95, 98, NT 2000, and XP
- Mac OS X (Intel)
- Linux/x86_64