cuidachao / h-blast

heterogeneous BLAST (H-BLAST), a fast parallel search tool for a heterogeneous computer that couples CPUs and GPUs, to accelerate BLASTX and BLASTP – basic tools of NCBI-BLAST.

Heterogeneous BLAST (H-BLAST)

Heterogeneous BLAST (H-BLAST) is a fast parallel search tool for a heterogeneous computer that couples CPUs and GPUs, to accelerate BLASTX and BLASTP - basic modules of NCBI-BLAST. H-BLAST employs a locally decoupled seed-extension algorithm to take advantages of GPUs, and offers a performance tuning mechanism for better efficiency among various CPUs and GPUs combinations. H-BLAST produces identical alignment results as NCBI-BLAST and its computational speed is much faster than that of NCBI-BLAST.

H-BLAST has been tested on different software and hardware settings. For example, CentOS Linux (v5.4, v6.0 and v7.1) and Ubuntu Linux (v11.04) were used as H-BLAST's benchmark software platforms; NVIDIA Tesla C2050 GPU card, NVIDIA Tesla K20/K40 GPU card and NVIDIA Geforce GTX 560 GPU card were used as H-BLAST's benchmark hardware platforms.

H-BLAST is free for academic and non-commercial use. The current implementation is based on NCBI-BLAST, and immigrates some code segments from GPU-BLAST.

Please cite the authors in any work or product based on this material: Weicai Ye, Ying Chen, Yongdong Zhang, Yuesheng Xu; H-BLAST: a fast protein sequence alignment toolkit on heterogeneous computers with GPUs. Bioinformatics 2017; 33 (8): 1130-1138. doi: 10.1093/bioinformatics/btw769.

For any questions and feedback about H-BLAST, contact [email protected] or [email protected].

Features:

• H-BLAST accelerates BLASTX and BLASTP - basic modules of NCBI-BLAST.
• H-BLAST can take advantage of multiply CPU cores and GPUs in a computer.
• H-BLAST offers a performance tuning mechanism for better efficiency among various CPUs and GPUs combinations, including manual and automatic tuning.
• The GPU memory demand for H-BLAST is low, that for on-the-fly results and alignment results are as less as 1/8 and 1/2 of those in GPU-BLAST, respectively. Even a Geforce GTX 560 GPU card with 1GB ram can be used for H-BLAST to accelerate BLASTX.
• H-BLAST v1.1 does not support PSI BLAST. The substitution matrix is fixed to the Blosum62 matrix.

Requirement:

• Nvidia GPU card with compute capability >= 2.0
• CUDA version >= 5.5.
• NCBI-BLAST v2.2.28+