an ultra-high-performance protein-protein docking for heterogeneous supercomputers

Requirements

• FFTW3 - http://www.fftw.org
  --enable-float flag must be specified when you compile FFTW3 (see also FAQ).
• OpenMPI - http://www.open-mpi.org (if you use MPI)
  
• CUDA Toolkit ver >= 5.0 - https://developer.nvidia.com/cuda-zone (if you use GPU)
  
• GPU Computing SDK code samples (CUDA SDK, same version as CUDA Toolkit) - https://developer.nvidia.com/cuda-zone (if you use GPU)

Installation

You can use MEGADOCK 4.0 on (a) GPU cluster, (b) CPU cluster, (c) GPU single node and (d) CPU single node. Please see the appropriate instructions below.

(a) GPU cluster (GPU, MPI & OpenMP hybrid parallelization)

Extract tarball contents

      $ tar xzf megadock-4.1.1.tgz
      $ cd megadock-4.1.1

Edit Makefile

    CUDA_INSTALL_PATH ?= your/cuda/toolkit/install/path # default: /usr/local/cuda
    CUDA_SAMPLES_PATH ?= your/cuda/sdk/install/path     # default: /usr/local/cuda/samples
    FFTW_INSTALL_PATH ?= your/fftw/library/install/path # default: /usr/local
    CPPCOMPILER       ?= g++                            # default: g++ (or icpc, others)
    MPICOMPILER       ?= mpicxx                         # default: mpicxx (or others)
    OPTIMIZATION      ?= -O3                            # 
    OMPFLAG           ?= -fopenmp                       # default: -fopenmp (g++), or -openmp, -qopenmp (intel)

Compile

      $ make

A binary file megadock-gpu-dp will be generated.

(b) CPU cluster (MPI & OpenMP hybrid parallelization)

Extract tarball contents

      $ tar xzf megadock-4.1.1.tgz
      $ cd megadock-4.1.1

Edit Makefile

    FFTW_INSTALL_PATH ?= your/fftw/library/install/path # default: /usr/local
    CPPCOMPILER       ?= g++                            # default: g++ (or icpc, others)
    MPICOMPILER       ?= mpicxx                         # default: mpicxx (or others)
    OPTIMIZATION      ?= -O3                            # 
    OMPFLAG           ?= -fopenmp                       # default: -fopenmp (g++), or -openmp, -qopenmp (intel)
    
    USE_GPU := 0 

Compile

      $ make

A binary file megadock-dp will be generated.

(c) GPU single node (GPU parallelization)

Extract tarball contents

      $ tar xzf megadock-4.1.1.tgz
      $ cd megadock-4.1.1

Edit Makefile

    CUDA_INSTALL_PATH ?= your/cuda/toolkit/install/path # default: /usr/local/cuda
    CUDA_SAMPLES_PATH ?= your/cuda/sdk/install/path     # default: /usr/local/cuda/samples
    FFTW_INSTALL_PATH ?= your/fftw/library/install/path # default: /usr/local
    CPPCOMPILER       ?= g++                            # default: g++ (or icpc, others)
    OPTIMIZATION      ?= -O3                            # 
    OMPFLAG           ?= -fopenmp                       # default: -fopenmp (g++), or -openmp, -qopenmp (intel)
    
    USE_MPI := 0 

Compile

      $ make

A binary file megadock-gpu will be generated.

(d) CPU single node (OpenMP thread parallelization)

Extract tarball contents

      $ tar xzf megadock-4.1.1.tgz
      $ cd megadock-4.1.1

Edit Makefile

    FFTW_INSTALL_PATH ?= your/fftw/library/install/path # default: /usr/local
    CPPCOMPILER       ?= g++                            # default: g++ (or icpc, others)
    OPTIMIZATION      ?= -O3                            # 
    OMPFLAG           ?= -fopenmp                       # default: -fopenmp (g++), or -openmp, -qopenmp (intel)

    USE_GPU := 0
    USE_MPI := 0

Compile

      $ make

A binary file megadock will be generated.