Limits...
Parallel implementation of 3D protein structure similarity searches using a GPU and the CUDA.

Mrozek D, Brożek M, Małysiak-Mrozek B - J Mol Model (2014)

Bottom Line: Graphics processing units (GPUs) and general purpose graphics processing units (GPGPUs) can perform many time-consuming and computationally demanding processes much more quickly than a classical CPU can.The GPU (GeForce GTX 560Ti: 384 cores, 2GB RAM) implementation of CASSERT ("GPU-CASSERT") parallelizes both alignment phases and yields an average 180-fold increase in speed over its CPU-based, single-core implementation on an Intel Xeon E5620 (2.40GHz, 4 cores).In this paper, we show that massive parallelization of the 3D structure similarity search process on many-core GPU devices can reduce the execution time of the process, allowing it to be performed in real time.

View Article: PubMed Central - PubMed

Affiliation: Institute of Informatics, Silesian University of Technology, Gliwice, Poland, dariusz.mrozek@polsl.pl.

ABSTRACT
Searching for similar 3D protein structures is one of the primary processes employed in the field of structural bioinformatics. However, the computational complexity of this process means that it is constantly necessary to search for new methods that can perform such a process faster and more efficiently. Finding molecular substructures that complex protein structures have in common is still a challenging task, especially when entire databases containing tens or even hundreds of thousands of protein structures must be scanned. Graphics processing units (GPUs) and general purpose graphics processing units (GPGPUs) can perform many time-consuming and computationally demanding processes much more quickly than a classical CPU can. In this paper, we describe the GPU-based implementation of the CASSERT algorithm for 3D protein structure similarity searching. This algorithm is based on the two-phase alignment of protein structures when matching fragments of the compared proteins. The GPU (GeForce GTX 560Ti: 384 cores, 2GB RAM) implementation of CASSERT ("GPU-CASSERT") parallelizes both alignment phases and yields an average 180-fold increase in speed over its CPU-based, single-core implementation on an Intel Xeon E5620 (2.40GHz, 4 cores). In this paper, we show that massive parallelization of the 3D structure similarity search process on many-core GPU devices can reduce the execution time of the process, allowing it to be performed in real time. GPU-CASSERT is available at: http://zti.polsl.pl/dmrozek/science/gpucassert/cassert.htm.

Show MeSH

Related in: MedlinePlus

Calculation of the similarity matrix S in the second phase of alignment. Molecular residue descriptors of the candidate database structure are (virtually) located along the vertical edge of the matrix and molecular residue descriptors of the query protein structure are located along the horizontal edge of the matrix. Calculations are performed in areas of size 4×4. Values of the cells in these areas are calculated according to the given order. Colors reflect the type of read/write operation that are required and the memory resources that are affected
© Copyright Policy - OpenAccess
Related In: Results  -  Collection


getmorefigures.php?uid=PMC3936136&req=5

Fig10: Calculation of the similarity matrix S in the second phase of alignment. Molecular residue descriptors of the candidate database structure are (virtually) located along the vertical edge of the matrix and molecular residue descriptors of the query protein structure are located along the horizontal edge of the matrix. Calculations are performed in areas of size 4×4. Values of the cells in these areas are calculated according to the given order. Colors reflect the type of read/write operation that are required and the memory resources that are affected

Mentions: Transfer of data packages to the device is performed in the same manner as in the first phase. Four streams are used for this purpose. After the first part of the data has been transferred to the GPU device, the high-resolution alignment procedure is initiated. Block threads perform parallel alignments of chains of molecular residue descriptors. Each block thread performs a pairwise alignment of the query protein vs. one candidate database protein. In order to limit the number of accesses to the global memory of the GPU device, the similarity matrix S is divided into rectangular areas of size 4 × 4. Calculations are performed area by-area, and row by row inside each area, from left to right, as shown in Fig. 10.Fig. 10


Parallel implementation of 3D protein structure similarity searches using a GPU and the CUDA.

Mrozek D, Brożek M, Małysiak-Mrozek B - J Mol Model (2014)

Calculation of the similarity matrix S in the second phase of alignment. Molecular residue descriptors of the candidate database structure are (virtually) located along the vertical edge of the matrix and molecular residue descriptors of the query protein structure are located along the horizontal edge of the matrix. Calculations are performed in areas of size 4×4. Values of the cells in these areas are calculated according to the given order. Colors reflect the type of read/write operation that are required and the memory resources that are affected
© Copyright Policy - OpenAccess
Related In: Results  -  Collection

Show All Figures
getmorefigures.php?uid=PMC3936136&req=5

Fig10: Calculation of the similarity matrix S in the second phase of alignment. Molecular residue descriptors of the candidate database structure are (virtually) located along the vertical edge of the matrix and molecular residue descriptors of the query protein structure are located along the horizontal edge of the matrix. Calculations are performed in areas of size 4×4. Values of the cells in these areas are calculated according to the given order. Colors reflect the type of read/write operation that are required and the memory resources that are affected
Mentions: Transfer of data packages to the device is performed in the same manner as in the first phase. Four streams are used for this purpose. After the first part of the data has been transferred to the GPU device, the high-resolution alignment procedure is initiated. Block threads perform parallel alignments of chains of molecular residue descriptors. Each block thread performs a pairwise alignment of the query protein vs. one candidate database protein. In order to limit the number of accesses to the global memory of the GPU device, the similarity matrix S is divided into rectangular areas of size 4 × 4. Calculations are performed area by-area, and row by row inside each area, from left to right, as shown in Fig. 10.Fig. 10

Bottom Line: Graphics processing units (GPUs) and general purpose graphics processing units (GPGPUs) can perform many time-consuming and computationally demanding processes much more quickly than a classical CPU can.The GPU (GeForce GTX 560Ti: 384 cores, 2GB RAM) implementation of CASSERT ("GPU-CASSERT") parallelizes both alignment phases and yields an average 180-fold increase in speed over its CPU-based, single-core implementation on an Intel Xeon E5620 (2.40GHz, 4 cores).In this paper, we show that massive parallelization of the 3D structure similarity search process on many-core GPU devices can reduce the execution time of the process, allowing it to be performed in real time.

View Article: PubMed Central - PubMed

Affiliation: Institute of Informatics, Silesian University of Technology, Gliwice, Poland, dariusz.mrozek@polsl.pl.

ABSTRACT
Searching for similar 3D protein structures is one of the primary processes employed in the field of structural bioinformatics. However, the computational complexity of this process means that it is constantly necessary to search for new methods that can perform such a process faster and more efficiently. Finding molecular substructures that complex protein structures have in common is still a challenging task, especially when entire databases containing tens or even hundreds of thousands of protein structures must be scanned. Graphics processing units (GPUs) and general purpose graphics processing units (GPGPUs) can perform many time-consuming and computationally demanding processes much more quickly than a classical CPU can. In this paper, we describe the GPU-based implementation of the CASSERT algorithm for 3D protein structure similarity searching. This algorithm is based on the two-phase alignment of protein structures when matching fragments of the compared proteins. The GPU (GeForce GTX 560Ti: 384 cores, 2GB RAM) implementation of CASSERT ("GPU-CASSERT") parallelizes both alignment phases and yields an average 180-fold increase in speed over its CPU-based, single-core implementation on an Intel Xeon E5620 (2.40GHz, 4 cores). In this paper, we show that massive parallelization of the 3D structure similarity search process on many-core GPU devices can reduce the execution time of the process, allowing it to be performed in real time. GPU-CASSERT is available at: http://zti.polsl.pl/dmrozek/science/gpucassert/cassert.htm.

Show MeSH
Related in: MedlinePlus