XPFS

• A New Parallel PROSITE Profile Search Algorithm on Xeon Phi.

Introduction

• XPFS is a high performance database searching method based on the Intel(R) MIC platform. We have used a two-level parallelization scheme: the thread level coarse-grained and VPU level fine-grained parallelism to implement our algorithm. We gain 3 times faster than PFSearchV3 on the Xeon Phi 7110p.

Installation and Usage

• Query profiles and database are needed to run XPFS. Profiles can be downloaded at [ftp://lausanne.isb-sib.ch/pub/software/unix/pftools/pftoolsV3/pftoolsV3.PROSITE.tar.gz]. And databse should be in FASTA format.

Run XPFS

• step 1 : Preproccess the database ( Generate database index file )

  ./XPFS -t 1 -T 0 -c <name of database index> <queryProfile> <database>

• step 2 : Do searching

  ./XPFS -t nCpus -T nMics -i <name of database index> <queryProfile> <database>

  nCPUs : number of cpu thread; nMICs : number of Xeon Phi cards (Currently only one card supported)

• for more details, you can run :

  ./XPFS -h

Reference

Authors and Contributors

• Weiguo Liu is the corresponding author.

Support or Contact

• If you have any questions, please contact: Weiguo,Liu ( weiguo.liu@sdu.edu.cn).