EGI-Engage:Data Plan

From EGIWiki
Jump to: navigation, search
EGI-Engage project: Main page WP1(NA1) WP3(JRA1) WP5(SA1) PMB Deliverables and Milestones Quality Plan Risk Plan Data Plan
Roles and
responsibilities
WP2(NA2) WP4(JRA2) WP6(SA2) AMB Software and services Metrics Project Office Procedures


Contents


Help and support: quality@egi.eu

Last update: March 2016

This page describe data management plan for the research data that will be generated within EGI-Engage. For each dataset, it describes the type of data and their origin, the related metadata standards, the approach to sharing and target groups, and the approach to archival and preservation.

Deliverable 2.4 Data management plan

This document will be further developed before the mid-term and final project reviews:

with more detailed information related to the discoverability, accessibility and exploitation of the data.

Rules

The Open Research Data Pilot applies to two types of data:


The obligations arising from the Grant Agreement of the projects are (see article 29.3):

Regarding the digital research data generated in the action (‘data’), the beneficiaries must:

  1. deposit in a research data repository and take measures to make it possible for third parties to access, mine, exploit, reproduce and disseminate — free of charge for any user — the following: the data, including associated metadata, needed to validate the results presented in scientific publications as soon as possible; other data, including associated metadata, as specified and within the deadlines laid down in the 'data management plan';
  2. provide information — via the repository — about tools and instruments at the disposal of the beneficiaries and necessary for validating the results (and — where possible — provide the tools and instruments themselves).

Note: As an exception, the beneficiaries do not have to ensure open access to specific parts of their research data if the achievement of the action's main objective, as described in Annex 1, would be jeopardised by making those specific parts of the research data openly accessible. In this case, the data management plan must contain the reasons for not giving access.

Datasets

Task Contact Data description Standards and metadata Data sharing Archiving and preservation
SA2.1/ SA2.2 Gergely Sipos (gergely.sipos@egi.eu)
  • Types of data: Feedback and requirements from existing and new EGI users are collected at training events and other types of face-to-face and electronic interactions. These data must be stored, managed, analysed and used efficiently because they represent high value for the EGI community to evolve its service portfolio.
  • Origin of data: collected from existing and potential users of EGI
  • Scale of data: few MB / year
The data is not in any standard format. Survey data - textual data, structured data (typically CSV or XLS) or graphics (usually survey summary or analysis)
  • Target groups: technology provider and service developer and provider teams who contribute to the EGI service portfolio
  • Scientific Impact: used for the further-development of IT services offered by the EGI Community. These services are often result of technological R&D and subject of publications in conference proceedings and peer-review journals
  • Approach to sharing: A public version of the collected requirements is going to be shared in the EGI-Engage milestones and deliverables. Some of the most important documents in this respect will be: M6.5 Joint training program for the second period (M15, May 2016), Intermediate and annual project reports (every 6 months), First milestones and deliverables of the ELIXIR, EPOS, BBMRI competence centres

Based on the nature of the data these can be:

SA2.3 Kimmo Mattila (kimmo.mattila@csc.fi) The use cases of Elixir Competence Center as well as  ELIXIR as infrastructure manage life science data  produced by life scientists. The decripition below refers to that data.

The Elixir Compentence Center itself is not creating new scientific data.
Due to that Elixir CC is not actively doing scientific data management.

  • Types of data: life science data; the management of genomics data: Marine metagenomics, Plant genomics and phenotype and Human sensitive data
  • Origin of data: produced and submitted by scientists. ELIXIR repositories collect, integrate and provide access to the data.
  • Scale of data: The biggest data collections in life sciences are in the order of petabytes (PB), however, it is likely that the ELIXIR CC will work with smaller data sets. A single whole human genome raw data is roughly 200 GB.
Some standards like the standard formats in the marine or the plain domain are still under development. Some of the standards for capturing and exchanging genomic data that might be used in the use cases are described in BioSharing [R3]. Part of the data may be stored to public data repositories (e.g. ENA) that have clearly defines metadata models.
  • Target groups: researchers interested to submit or use Metagenomics, Plant and Human data.
  • Scientific Impact: scientific discoveries such as comparative environmental metagenomic analyses or finding genes related to a disease
  • Approach to sharing: ELIXIR promotes open data access, but naturally human data might be sensitive therefore requires authorised access.
Services for archiving and preservation within ELIXIR are listed in https://www.elixir-europe.org/services.
SA2.4

Petr Holub (petr.holub@bbmri-eric.eu)

Deals primarily with human-related data, most of which can be considered privacy-sensitive.
  • Types of data: privacy-sensitive human data, limited data sets of non-human data.
  • Origin of data: health care (for patients participating in research), biobanks (collection of information for non-patient research participants, as well as analysis of samples), scientists (analysis of samples and data and return of resulting data back to the biobanks).
  • Scale of data: depending on the specific data types collected in the given setting, but various omics data and large 2D/3D imaging data can be in order of more than 1TB per research participant.
Standards are related to the origin of data: for clinical biobanks, it is related to the healthcare standards in the given country (not homogeneous across Europe); for population biobanks, it is subject to various ongoing efforts to make data structures more standardized and interoperable (e.g., ISO TC276 Working Group 5).
  • Target groups: researchers medical and biomedical research.
  • Scientific Impact: health care improvement, public health improvement.
  • Approach to sharing: FAIR (findable, accessible, interoperable, reusable) access, while complying with regulatory frameworks related to privacy-sensitive data.
Part of core business of biobanks and hence BBMRI-ERIC.
SA2.5 Alexandre Bonvin (a.m.j.j.bonvin@uu.nl)
  • Types of data: There is research data involved in the activity, but this is not produced with EGI-Engage resources, but from other EU projects. The types of data produced by those other projects are experimental NMR, Xray, SAXS and cryo-EM data.
  • Origin of data: Biological samples (owned by the end users of the facilities).
  • Scale of data:
The end results are typically deposited into public databases like the PDB or EMDB for cryo-EM data.
  • Target groups: The raw data are usually so complex that they are only of use to expert users in structural biology that have been trained in a specific technique. The processed and derived data typically deposited in public databases are of use to researchers in life sciences in general and for biotech and pharmaceutical companies.
  • Scientific Impact: This research data can underpin scientific publications.
  • Approach to sharing: Data are shared via databases (e.g. again PDB, EMDB), with possibly an embargo period until publication. Other datasets (e.g. the results of computations) can be shared via EUDAT or other repositories like SBGRID for structural biology. For such an example see: https://data.sbgrid.org/dataset/131/
From a university perspective, data are to be kept for 10 years. Currently, there is no proper archiving mechanism in place at the particular site (Utrecht University). At the moment, policies and services rely on what is provided by the database service providers where data are deposited.
SA2.6 Davor Davidović (davor.davidovic@irb.hr)
  • Types of data: the centre will generate/collect data that come from the research activities in the fields of Arts and Humanities. Common types of research data generated and collected in A&H are books, letters, emails, paintings, photographs, manuscripts, various digital collections, audio/video materials, etc. However, in the research activities related to EGI-Engage, the focus is on digitised data, i.e. the information/data stored in different digital formats, such as plain files (text, photo, audio and video), metadata, collections, and annotations.
  • Origin of data: The digitised data used in these research activities originates from the physical objects/artefacts used in the research activities connected to A&H, for example, books, audio and video materials, paintings, archaeological artefacts, etc. that can be found in museums, libraries, etc. However, the focus is on existing digitised collections of these physical artefacts that are generated, operated and managed by the members of the DARIAH community. Thus, the main sources of data for this related research are those digitised data provided by various DARIAH members. Some of DARIAH members already operate their own digital repositories, which will be used as a data source.
  • Scale of data: It is hard to estimate the scale of the research data because of a large number of different sources and the amount of information that is stored.
The community does not promote any specific metadata standard. The adopted metadata formats vary from case to case. Also, there is no recommendation about any long-term preservation format and thus no domain-specific data format is used or recommended. Thus, an individual approach for each use case is required.
  • Target groups: The data collected within the DARIAH Competence Centre will be useful primarily to the members of the DARIAH community. In addition, it is believed that the wider audience having strong interests in exploiting A&H data can benefit in using these data.
  • Scientific Impact: This research data can underpin scientific publications.
  • Approach to sharing: For now, no further information on how data will be shared and accessed is known. , the majority of data are stored and shared via various data repositories that can be widely accessed. The repositories are mostly institutional (i.e. DARIAH member institutions such as computing/storage centres or research organisations).
The implementation of the repositories, safe guarantee, number of copies, etc. is on individual data/repository providers. The plan is to implement several digital repositories for a specific DARIAH use cases (e.g. Bavarian dialects) using gLibrary framework that allows storing the data on different storage elements (local, grid and cloud storage elements).
SA2.7 Jesus Marco de Lucas (marco@ifca.unican.es)
  • Types of data: The LifeWatch competence centre will generate/collect mainly test datasets as part of larger datasets, to analyse the LifeWatch-EGI Competence Centre framework. For example, one month of data collected at a water reservoir, or six different simulation outcomes related to it.
  • Origin of data: Instruments in the water reservoir.
  • Scale of data: GB of data in a database that can be exported in the CSV file format.
As data formats for the data gathered

in the water reservoir we will use basically text-based formats like CSV. However, other data product will be generated using results of models performed over these input data and formats like NetCDF or HDF5 will be used. For sharing these datasets, the metadata standard used will be WaterML. WaterML is a OGC (Open Geospacial Consortium) standard information model for the representation of water observations data. In order to guarantee reproducibility of the experiments, the goal is to set up an ontology including WaterML attribute that establish relationships between the different components of the case study: instrumentation, datasets, software, models, etc.

  • Target groups: The data can be interesting for other research teams that make similar analysis at other water reservoirs.
  • Scientific Impact: The data can potentially underpin scientific publications.
  • Approach to sharing: The embargo period is usually two years as the data is exploited by an SME. The datasets released are those who are connected to scientific publications or referenced in public management reports. The repository is located at the IFCA data centre and freely accessible via web [R5], but registration is needed.
Copies are kept in WORM tapes, and in a separate server (400 km away) of the company. Main repository uses RAID technology and has not lost any data in the last 10 years. The data are automatically synchronised across the servers.
SA2.8 Ingemar Häggström (ingemar.haggstrom@eiscat.se)
  • Types of data: Development of value-added products (e.g. processes, combined data, plots).
  • Origin of data: EISCAT Incoherent Scatter radar low-level data.
  • Scale of data: A few TB/year will be produced within EGI-Engage. EISCAT data are of a larger order of magnitude.
A mixture of standards are adopted depending on type. For long-term preservation, the format hdf5 will be used.
  • Target groups: Mainly space and environmental researchers.
  • Scientific Impact: This research data can underpin scientific publications.
  • Approach to sharing: Current value-added products are open to all from day zero, but low-level data is not. Discussions on the new products are still on going.
Data are stored on a few e-Infrastructures, mirrored and synchronised. There are two levels of storage: a large short-term, and a reduced long-term.
SA2.9
Daniele Bailo (daniele.bailo@ingv.it)

  • Types of data: Seismological waveforms.
  • Origin of data: World-wide data archives providing FDSN interfaces for raw and earthquakes parametric data, (EIDA - IRIS - USGS - NCEC) and synthetics produced and postprocessed in HPC and local resources
  • Scale of data: Real streams form FDSN, including IRIS - US, are pre-staged on-demand. These are only limited by specific authorisation policies and mechanisms which are not yet fully hadled in the curren system. Therefore size of pre-staged raw data and processed synthetics depends from usage
FDSN standards for disseminating data. Data format : SEED and mSEED. Also: VTK, W3C-PROV, QuakeML, mp4, KML for the specific purpose each standard was created for.
  • Target groups: Mainly environmental researchers.
  • Scientific Impact: This research data can underpin scientific publications and development of pan european computational earth science e-Infrastructure
  • Approach to sharing: Produced data can be shared according to users' requirements. Authorization policies can be configured by users via the generic iRODS authorisation system and its GUI client. Default setting consider all data like 'private'. Metadata, provenance and lineage are currently publicly available.
Synthetic and pre-staged raw data are handled and stored within a local federation and data management system, which preserves lineage and provenance information. Original raw data is stored across the nodes of the FDSN service providers.
SA2.10 Eric Yen (Eric.Yen@twgrid.org)
  • Types of data: There are two main types of data:
    • Observation data from tidal gauge, weather stations, rainfall, radar data, satellite data and images, bathymetry, historical records of earthquake and tsunami, etc.
    • Waveform at any target site, potential source of a historical tsunami event, changes of rainfall, wind field and path of typhoon or any special weather event, dispersion path of aerosol or volcano ashes, are the primary simulation results.
  • Origin of data: Government agency of weather, earthquake, tsunami, and volcano; or research institutes that own the data needed by the CC.
  • Scale of data: Data scale of the whole collection and generated data would be few TB to 10s of TB. Variation is possible due to the resolution of the generated output.
The ISO 19156 standard for Observation and Measurement data model was selected. For weather and climate data, the centre will also comply with the Climate and Forecast convention (CF) (e.g. NetCDF). Both of these specifications are included in the new metadata model called ADAGUC Data format standard.
  • Target groups: The data can potentially underpin scientific publications. Scientists of tsunami, earthquake, volcano, weather, and climate changes; scientists, policy makers of disaster mitigation strategy and studies.
  • Scientific Impact: The data can support new discoveries such as the sources and characteristics of potential tsunami sources or new ways of hazards simulation and analysis. The data can also support new modelling schemes and the change processes of climate and disaster events.
  • Approach to sharing: Almost every government has strict regulation for announcement of weather and natural hazards, so the centre is focusing on research instead of releasing results to the public. Moreover, sharing is still up to the clearance of right for dissemination from the original agency. At least during the project years, the data collected or generated would be shared in a restricted way and for academic purposes only.
The data will be organised and managed in a repository over the distributed infrastructure. The CC plans to have no less than three copies of the data set at different sites. Academia Sinica (Taiwan) is in charge of the long-term data preservation.
Personal tools
Namespaces
Variants
Actions
Navigation
Toolbox
Print/export