MetaHipMer Version 2.0

News

Oct 3, 2023: The latest release is now 2.2.0.0. This largely includes a suite of bug fixes for improved performance and stability on GPU devices and some limited support for new slingshot hardware. There is an experimental docker image of MHM2 that should work to run on a single server at: https://hub.docker.com/repository/docker/robegan21/mhm2/general

January 31, 2023: the latest release is now 2.1.0.2. It offers support for both Nvidia and AMD GPUs, and features numerous performance improvements and bug fixes. However, we always recommend installing by pulling from the latest on the master branch. We endeavour to keep that branch stable, and it incorporates all the latest updates and bug-fixes that may not be in the latest official release.

July 27, 2022: a bug fix patch to support building with newer compilers was released 2.1.0.1.

January 24, 2022: a new release, version 2.1.0, is available. Several stages now can take advantage of Nvidia GPUs, with significant performance improvements on systems with GPUs (e.g. up to 4.5x at scale on the Summit supercomputer). What's new in this release can be seen here. For details on how to install and run MHM2, please consult the User guide.

Overview

MetaHipMer is a de novo metagenome short-read assembler. Version 2 (MHM2) is written entirely in UPC++, CUDA and HIP, and runs efficiently on both single servers and on multinode supercomputers, where it can scale up to coassemble terabase-sized metagenomes. More information about MetaHipMer can be found in:

• E. Georganas et al., "Extreme Scale De Novo Metagenome Assembly," SC18: International Conference for High Performance Computing, Networking, Storage and Analysis, Dallas, TX, USA, 2018, pp. 122-13.
• Hofmeyr, S., Egan, R., Georganas, E. et al. Terabase-scale metagenome coassembly with MetaHipMer. Sci Rep 10, 10689 (2020).

Quality of MetaHipMer assemblies is comparable with other leading metagenome assemblers, as noted in the results of the CAMI2 competition, where MetaHipMer was rated first for quality in two out of three datasets, and second in the third dataset:

• F. Meyer et al., "Critical Assessment of Metagenome Interpretation: the second round of challenges", Nature Methods volume 19, pages429–440 (2022)

Acknowledgements

MHM2 is developed and maintained by the Exabiome Project at Lawrence Berkeley National Laboratory, and is supported by the Exascale Computing Project (17-SC-20-SC), a collaborative effort of the U.S. Department of Energy Office of Science and the National Nuclear Security Administration.

Previous releases

Releases of MetaHipMer prior to Version 2 can be found at the HipMer download page.

Contact information

• MHM2 Issue tracker - bug reporting, feature requests, etc.
• Email