High memory computer hardware for metagenomic analysis
Lead Research Organisation:
University of Manchester
Department Name: Life Sciences
Abstract
Abstracts are not currently available in GtR for all funded research. This is normally because the abstract was not required at the time of proposal submission, but may be because it included sensitive information such as personal details.
Planned Impact
N/A
Organisations
| Description | The asset, now known as Hydra, was sourced, purchased and commissioned as planned, coming on-stream for users in spring 2014. At that stage the machine had 4 compute nodes, each supplied with 0.5Tb RAM. Due to improvements in the campus network it was possible to install this in a central location, allowing on-going maintenance and systems administration by the University-level research-computing team. Numerous applications have been made available to users of Hydra, of which a number of the most heavily used have had their usage on Hydra specifically documented (see http://ri.itservices.manchester.ac.uk/hydra-apps/software/a-z/ ). This includes both general tools like R and more specific tools like the phylogenetic tree builder RAxML. An important element of the original application was the creation of a local Galaxy server on Hydra, to enable access to users without command-line skills. This has been done in an installation that includes 60 different applications, including various meta-genomic tools, notably Mothur, set up. Hydra is being used by an increasingly large number of researchers within the research groups named on the original application, with peak usage reaching over 80,000 CPU hours a month. Development of the Hydra Galaxy server has been associated with user needs, which have been more episodic. One of the advantages of a central location and administration for Hydra is that it can be more readily used alongside other computing resources in the University of Manchester. This has also enabled others to see the benefits of such high memory hardware, meaning that the original asset has been built along-side to create a better and more widely shared facility. In summer 2014, three individual research groups and a Wellcome Trust Institutional Strategic Support Fund contributed to more than doubling the number of compute nodes available on Hydra and upgrading its networking to infiniBand. The University of Manchester contributed £250,000 of high memory hardware, increasing both nodes and storage accessible to Hydra as well as some extremely high memory nodes of 2Tb RAM or more. Each of these extensions extends the user-base beyond the original metagenomics community and improves the resources available to all users, thereby maximising the value achieved by the original asset for the original users. |
| Exploitation Route | The rapid up-take of usage demonstrates the research need being fulfilled. At the same time there has been a clear increase in related research in Manchester and Hydra has been and will continue to be used for training, increasing skills in computational biology. Hydra has also enabled cutting edge development of metagenomic approaches. Notably, the British Ecological Society's Plants, Soils, Ecosystems special interest group held a metagenomic sequencing meta-analysis workshop in Manchester in May 2015 led by one of the PIs named on the original application and the facility has been cited in several successful grants that use metagenomics, including some funded NERC. |
| Sectors | Agriculture, Food and Drink,Digital/Communication/Information Technologies (including Software),Education,Environment |
| Description | The equipment is being used by staff in the University for a range of applications. |
| First Year Of Impact | 2014 |