MPI-HMMER is an open source MPI implementation of the HMMER protein sequence analysis suite. The main search algorithms, hmmpfam and hmmsearch, have been ported to MPI in order to provide high throughput HMMER searches on modern computational clusters. We improve on HMMER through sophisticated I/O, a self-contained coordinator/worker model, and the easy inclusion of accelerated architectures. This results in better scalability while still maintaining the familiar user interface.

MPI-HMMER Features:

- Improved database chunking strategy
- Portable across any POSIX operating system
- MPI implementation independent
- Vastly reduced computation times
- Improved query throughput
Output nearly identical to standard HMMER

News: Announcing the first release of GPU-HMMER

The MPI-HMMER team is pleased to announce the release of our GPU-based hmmsearch tool. This initial release supports a single NVIDIA-CUDA capable GPU. It has been confirmed to be compatible with 8800 GTX Ultra and GTX 200 series GPUs, though it may also work on earlier CUDA-compatible cards as well. In the coming weeks we plan to release our multi-GPU implementation as well as MPI support for our GPU work.

Upcoming Releases:

In the coming months we plan to release several major enhancements to the MPI-HMMER package. We currently have several research groups implementing and porting MPI-HMMER to new platforms, and introducing support for additional features. We plan to introduce support for parallel IO to support users with high-end parallel file systems.

Separately, researchers are porting the core of MPI-HMMER to the Cell BE architecture with full support for SPEs. This will allow the integration of Cell compute nodes along with traditional general purpose processors to provide even greater compute power.

In the long term, MPI-HMMER is undergoing a major architectural overhaul designed to further improve the scalability of the MPI-HMMER searches for large (massively) parallel systems. More information regarding MPI-HMMER's architectural improvements will be made available as it approaches a usable, more stable, solution.

For researchers who find MPI-HMMER a useful tool, we would appreciate being cited for our work. Please use the following to cite MPI-HMMER:

J.P. Walters, B. Qudah, and V. Chaudhary . Accelerating the HMMER Sequence Analysis Suite Using Conventional Processors . In Proceedings of AINA 2006, Vienna Austria.

J. Landman, J. Ray, and J.P. Walters . Accelerating HMMer searches on Opteron processors with minimally invasive recoding. In Proceedings of HiPCOMB 2006, Vienna Austria.