Welcome to pblat-cluster

This program is for computing cluster with MPI support. For single computer with multiple core processors, please refer to pblat.

When the query file format is fasta, you can specify many processes in a cluster to process it. Processes running in the same node use shared memory. Thus pblat-cluster works in hybrid computing mode. This will minimize the memory usage per node. It can reduce run time linearly. This program is useful when you blat a big query file to a huge reference like human whole genome sequence.

The program is based on the original blat program which was written by Jim Kent.

pblat-cluster can run on Linux clusters with MPI support.


To compile the source code, simply enter the source code directory in terminal and issue the "make" command. When the compiling finished, the executable pblat-cluster will locate in the same directory. Then it can be moved to where you want.

$ make

This will use mpicc to compile the codes by default. You can specify other MPI compilers installed in your cluster. For example, using Intel MPI compiler:

$ make CC=mpiicc


Two ways to run pblat-cluster in a cluster:

without PBS

$ mpirun -n <N> pblat-cluster database query output.psl

with PBS

You can write a bash script and submit to PBS using qsub/sbatch.

#PBS -N pblat
#PBS -l nodes=32:ppn=4

cd workingdir

mpirun pblat-cluster genome.fa reads.fa out.psl
#SBATCH -J pblat
#SBATCH -n 4

cd workingdir

mpirun pblat-cluster genome.fa reads.fa out.psl


pblat is modified from blat, the licence is the same as blat. The source code and executables are freely available for academic, nonprofit and personal use. Commercial licensing information is available on the Kent Informatics website.

Authors and Contributors

Copyright (C) 2012 - 2015 Wang Meng

Contact me: wangm@mail.cbi.pku.edu.cn