1 of 68

Open Science in the project lifecycle - access & reuse

Elin Kronander

Open Science in the Swedish context, 2025-05, Stockholm

2 of 68

Repositories are key for share and reuse

Re-colored data Life Cycle by RDMkit used under CC-BY

3 of 68

Reuse - open is not enough

4 of 68

Reuse - open is not enough

What do we need in order to be able to reuse a research output?

5 of 68

Reuse - open is not enough

What do we need in order to be able to reuse a research output?

That the object exists/where to look for it
Detailed information about the object
How the object can be accessed
How the object could/should be used
Who and how has the right to use it
…

6 of 68

Reuse - open is not enough

What do we need in order to be able to reuse a research output?

That the object exists/where to look for it - Repositories
Detailed information about the object - Metadata
How the object can be accessed - Persistent Identifiers
How the object could/should be used - Metadata
Who and how has the right to use it - Licenses
…

7 of 68

Reuse - open is not enough

What do we need in order to be able to reuse a research output?

8 of 68

Screenshot from the SciLifeLab Data Repository

https://doi.org/10.17044/scilifelab.24514675.v1

9 of 68

Types of repositories

Domain/type-specific:

Best choice - long-term plan, typically free, maximum reach
E.g. European Nucleotide Archive, European Genome Phenome Archive, ArrayExpress, PRIDE

General purpose:

Second best - long-term plan, might cost (now or in future), good reach but less specific in metadata → more difficult for future users to judge if a dataset will be useful
E.g. Zenodo, (SciLifeLab) Figshare, SND Doris, Dryad

In-house/institutional

For archive/backup purpose mainly, might cost, limited reach unless also published in a data catalogue

Domain-specific

General �purpose

In-house

There are different types of repositories:

Domain specific repositories:

These are repositories dedicated to specific scientific domains (e.g., genomics, physics, ecology).
Best choice if there is a suitable one for your data type. They have long-term plan for sustainability, they are typically free of charge, and has maximum reach in your research community.
E.g. European Nucleotide Archive, ArrayExpress

General purpose repositories:

These repositories are designed to accept datasets from any field of research. They are more generalized, meaning they accept a broader range of data types.
If there is no domain specific repository, a general purpose repository is the best choice. They also have long-term plan for sustainability, but might cost (now or in future), and they do have good reach. However, the metadata is less specific which means it is more difficult for future users to judge if a dataset will be useful to them or not.
E.g. Zenodo, Figshare, Dryad
Some general purpose repositories have metadata profiles

In house/institutional repositories:

These are repositories that are run by specific institutions (universities, research centers) primarily for archiving and backup purposes
This type of repository is for archive/backup purpose mainly, since it has limited reach outside the institution, they are typically not ‘googlable’, unless also published in a public (indexed) data catalogue. Also, it might be associated with a cost.

Background:

Domain-specific repositories

This type of repository focuses on specific data types and is typically the best choice if you can find one that is suitable for your research data. It will reach your research community, so that others working in your field can find and reuse your data, and incorporates metadata standards in order to make the data as widely useful as possible. The repositories usually have long-term sustainability plan, i.e. they will be available for a long time, and are typically free of charge.

European Nucleotide Archive (ENA) - for genomic sequence data (non-human)
European Genome Phenome Archive (EGA) - for human genomic sequence data
ArrayExpress - for gene expression data
PRIDE - for proteomics data

It could also be a repository for a specific output such as

General purpose repositories

From time to time you will most likely come across situations when there is no suitable domain-specific repository for your data type, for example if you have a new data type. Another situation you might find yourself in is that you have done registry-based research with human data, and you are not allowed to share this data but still want to publish the methodology openly. Making a metadata record in a general purpose repository will then allow others to easily find it, without violating the agreements with the registry holders.

General purpose repositories usually accepts anything and everything related to research, i.e. they are also useful for other purposes, besides publishing research data, e.g. posters and presentations can be made publicly available and obtain a persistent identifier (DOI).

This type of repository is typically indexed, so you can find its content via search engines, and thus it has good reach. However, since the repository accepts many data types, the metadata will be less specific (no or very high level metadata standards), with the result that is more difficult for future users to judge if a dataset will be useful for them. The repositories usually have long-term sustainability plan, i.e. they will be available for a long time, but might cost (now or in future).

In-house/institutional repositories

Many institutions offer local repositories, and might even require that affiliated researchers put their research outputs there. However, unless the repository is indexed (so that search engines can find its content) this type of repository is for archive/backup purpose mainly.

You can also choose to create and host an in-house repository yourself, but that put a lot of responsibility on your shoulders. For how long will you be able to sustain it? It also requires considerable labour in order to make the repository FAIR, and without that the repository will have limited reach unless you also publish in a data catalogue.

10 of 68

Data catalogues and registries

Output specific registries

WorkflowHub - a registry for describing, sharing and publishing scientific computational workflows

Protocols.io a platform for organising, describing and sharing methods

ELIXIR TeSS - a registry for training materials

Data repositories

FAIRsharing.org/databases - catalogue of many repositories, with possibility to filter on e.g. domain

Scientific Data Repository Guidance - publisher’s recommendation

re3data.org - registry of research data repositories

Swedish data

researchdata.se - a national web portal where you can find, share, and reuse research data from a wide range of disciplines

dataportalen.se - a national web portal where you can find, share, and reuse data beyond research data

11 of 68

Restricted access repository: FEGA Sweden

National repository for storing and sharing personally identifiable information from Swedish biomedical research projects.

fega.nbis.se

12 of 68

FAIR principles

FAIR principles illustration. SND/Svensk nationell datatjänst. CC-BY 4.0

13 of 68

The FAIR data guiding principles

To be Findable:

F1. (meta)data are assigned a globally unique and persistent identifier

F2. data are described with rich metadata (defined by R1 below)

F3. metadata clearly and explicitly include the identifier of the data it describes

F4. (meta)data are registered or indexed in a searchable resource

To be Accessible:

A1. (meta)data are retrievable by their identifier using a standardized communications protocol

A1.1 the protocol is open, free, and universally implementable

A1.2 the protocol allows for an authentication and authorization procedure, where necessary

A2. metadata are accessible, even when the data are no longer available

To be Interoperable:

I1. (meta)data use a formal, accessible, shared, and broadly applicable language for knowledge representation.

I2. (meta)data use vocabularies that follow FAIR principles

I3. (meta)data include qualified references to other (meta)data

To be Reusable:

R1. meta(data) are richly described with a plurality of accurate and relevant attributes

R1.1. (meta)data are released with a clear and accessible data usage license

R1.2. (meta)data are associated with detailed provenance

R1.3. (meta)data meet domain-relevant community standards

14 of 68

FAIR does not mean open

FAIR and OPEN image from CESSDA Data Archiving Guide version 4.0. CESSDA Training Team (2025). CC-BY 4.0

15 of 68

Metadata

Open Science in the project lifecycle - access & reuse

16 of 68

Metadata

The data about the data (or anything really)

“One person’s metadata, is another person’s data”

17 of 68

Impactful metadata

Human readable
Standardised
Machine readable

18 of 68

Exercise instructions

In groups:

Examine the packaging of the food item
List the information you find

Think broadly

Consider why each piece of information is included in the packaging

19 of 68

Exercise reflection

Share 1-2 interesting pieces of information you found on the packaging
How would these elements relate to research metadata?
Going back to the previous slide, could you identify something that could be described as standardised metadata or machine readable metadata?

20 of 68

Elements relating to research metadata

Product Name → Research Title
Ingredients → Methods/Materials
Best Before Date → Creation Date/Version
Manufacturer → Author/Affiliation

21 of 68

Human readable

Machine readable

Standardised

22 of 68

Metadata at different levels

?

Source: Openclipart

?

.fastq

Source: Openclipart

?

Source: Publicdomainpictures

This slide, including all images, are in the public domain and free to reuse.

23 of 68

Metadata in the Data Life Cycle

Determine what metadata is needed for doing the project and sharing data

Collecting metadata

Describing

datasets

Using metadata to perform analysis

Describing datasets

Understanding datasets

Organising metadata in appropriate format(s)

RDM life cycle, by ELIXIR RDMkit, CC-BY

24 of 68

The beauty of formalised metadata

Describe the same thing! (i.e. Contextual terms)

Use the same words! (i.e. Recommended terminologies)

With good levels of detail! (i.e. Metadata templates)

Ok, but what is that?

Source: Openclipart

Repositories are your friends!

This slide, including all images, are in the public domain and free to reuse.

The answer is formalised metadata, to apply universal standards to your descriptions making it understandable to everyone, everywhere

[ click ] By using standard ways of describing how the samples were collected, and how the experiments were conducted, we can make sure that we all:

Describe the same thing - by using contextual terms that are universally recognised and specific for your field of research
Use the same words - by using terminologies that are agreed upon within the research community
Supply good levels of detail - by using metadata templates that ensure a rich level of information

And how can you get access to these community standards, you might ask?

The answer is [click] by using repositories!

Repositories that are tailored to specific data types, such as sequence data, have implemented appropriate standards on how the data should be described

25 of 68

Controlled Vocabularies & Ontologies

Controlled Vocabularies

A predefined, organised list of authorised terms
Ensures consistency in data description
Examples: Species taxonomies, Medical Subject Headings (MESH)

Ontologies

A structured framework showing relationships between concepts/terms
Supports complex data integration and interoperability

Benefits as Metadata Descriptors

Improved Data Consistency: Reduces ambiguity and enhances searchability
Enhanced Interoperability: Facilitates data sharing across systems
Efficient Data Management: Streamlines classification and retrieval processes

This slide, including all images, are in the public domain and free to reuse.

26 of 68

Metadata transformation

A simple example:

	Place
	Strängnäs

One location

=

One metadata value

Does anyone need to know more?

This slide, including all images, are in the public domain and free to reuse.

27 of 68

Metadata transformation

Geographic location (country and/or sea)	Geographic location (region and locality)	Geographic location (latitude)	Geographic location (longitude)
Sweden	Strängnäs	59.29	17.12

Yes!

More detail let others (and your future self) know what you have done

Rich metadata is the key to scientific reliability!

This slide, including all images, are in the public domain and free to reuse.

28 of 68

Persistent Identifiers

Open Science in the project lifecycle - access & reuse

This presentation originates from session 8 of the The ELIXIR FAIR training material by Design course. DOI:10.5281/zenodo.13773159

29 of 68

What is a unique persistent identifier (PID)

Identify

Verify

Locate

So to start with, what is a persistent identifier?

A persistent identifier (PID) is a long-lasting reference that uniquely tags a resource, for example a dataset, and provides the information required to reliably identify, verify and locate your resource eliminating many misunderstandings. (A PID may also be connected to a set of metadata which describes a digital or non-digital resource rather than to the item itself.) A PID ensures that access to the digital object is persistent. PIDs avoid broken links and difficulties locating a digital object (e.g. a dataset, a publication), even if its web address (URL) changes. A central registry ensures that following the PID will point you to the digital object’s current location.

�In order to make this a bit more concrete I’ll give you an example.

30 of 68

JANE DOE, HIGH ST 5

NORWAY

JANE DOE, MILL RD 1 �SWEDEN

JANE SMITH, CASHEW WAY

GERMANY

JANE

Let’s imagine that you are going to be invited to a high school reunion in the hometown of your youth. However, you've moved several times in your life. With each move, your physical address changes. You might also have changed your name at some point in your life. When the organising committee wants to send you the invitation, they need your current name and address but since you both changed your name and moved your invite will be lost. �But what if there was a magical mailbox that stayed the same no matter where you moved? You could give out this "magical mailbox" address once, and anyone could send you letters to that address, which would then automatically find its way to wherever you live now. Even if you move across the globe, your mail reaches you through this magical, never-changing mailbox.

A Persistent Identifier (PID) acts like this magical mailbox for digital information. Let's say there's a specific research paper online. It gets a PID, a unique digital "address" that never changes, even if the paper moves to different locations on the internet (like from one database to another). No matter how the internet changes or where the document goes, you can use this PID to find the paper, just like how your letter finds you through your unchanging, magical mailbox.

The primary purpose of the PID is to provide the information required to reliably identify, verify and locate the resource it is connected with. In order to do so, the PIDs must comply with a few rules:

31 of 68

Features of PIDs

Globally unique

It should comply with a controlled syntax to avoid clashes

Persistent

It should be maintained for a long period of time. The syntax used for the identifier should also be persistent

Resolvable

It should allow both human and machine users to access the resource

Globally unique:

To enable global uniqueness, a PID should comply with a controlled syntax to avoid clashes, for instance, by having namespaces that are governed by clearly defined authorities, ensuring that there are no two identical identifiers that point to different digital objects

Persistent:

The identifier, and the object to which it points, should be maintained for a long period of time. The syntax used for the identifier should be also persistent. A PID ensures that access to the digital object is persistent. PIDs avoid broken links and difficulties locating a digital object (e.g. a dataset, a publication), even if its web address (URL) changes. A central registry ensures that following the PID will point you to the digital object’s current location.�

Resolvable:�The identifier allows both human and machine users to access the resource or information

32 of 68

Are you familiar with some persistent identifiers?

33 of 68

Digital Object Identifier - DOI

Illustration from the DOI Foundations website�https://www.doi.org/the-identifier/what-is-a-doi/

An example of a persistent identifier that you might be familiar with is the DOI Digital Object Identifier (DOI) (doi.org) It is one of the most common PIDs used by public repositories. A DOI, like any other PID, is a long-lasting reference that uniquely tags a resource. While the identifier itself is digital the object it is identifying can be of any kind, Physical, digital or abstract. DOIs are and are actionable, meaning that you can resolve it using the web browser and be taken to a web page with the listed digital object and its metadata. Actual access to the digital object from this page might be restricted since a PID may be connected to a set of metadata describing an item rather than to the item itself. DOIs are coupled with metadata that can be modified over time and to keep track of the locations and characteristics of the objects they identify.

Designed to be used by humans as well as machines, DOIs identify objects persistently. They allow things to be uniquely identified and accessed reliably. You know what you have, where it is, and others can track it too. A DOI is a unique number made up of a prefix and a suffix separated by a forward slash. This is an example of one: 10.1000/182. It is resolvable using our proxy server by displaying it as a link: https://doi.org/10.1000/182.�Foundation and RAs The DOI Foundation is the governance body of the DOI System and is composed of registration agencies that provide services to their respective communities (people or organisations who need to identify and track the things that matter to them.) Their work involves allocating DOI prefixes, registering DOI names, and providing a metadata schema associated with each DOI record. How does this work?

34 of 68

Digital Object Identifier - DOI

Illustration from The DOI Handbook April 2023 (https://doi.org/10.1000/182 identifies the latest current version of the handbook)

This figure illustrates the basic principle of the DOI System. Let’s start in the middle. �Any entity (digital, physical, or abstract) can be identified by a global unique and persistent identifier called a DOI name. The DOI name can be resolved to a resource, such as a web or internet resource for example a Zenodo record, with metadata describing the entity, a landing page with access to further resources, etc. The resource could also be an instance of the entity it represents. For example, a DOI name representing an article resolves to the web address of the HTML file version of the article.

Metadata must be assigned to each DOI name to describe the entity represented by the DOI name. Metadata interoperability is ensured through basic principles outlined by the DOI Foundation. Metadata is used to provide services to the users: it can be displayed to users to enrich an information resource; it can be used by users to search for a DOI name; etc.

35 of 68

Other examples of PIDs

ORCID - Open Researcher and Contributor ID

persistent identifiers for researchers
takes homonymy into account
add aliases to your profile if your name changes
ORCID stays the same when affiliation changes

The Research Organization Registry

persistent identifiers for research organizations

ORCID, which stands for Open Researcher and Contributor ID, is a global, non-profit organisation which provides a unique and persistent identifier free of charge to researchers. Especially if you have a common name, you’ll know how important it is to distinguish homonyms! It is extremely useful to be correctly identified, worldwide. ORCID takes homonymy into account, and the system also allows you to add aliases to your profile in the event that your name changes, making sure that it will be tracked back to you. Another benefit is that your ORCID will stay the same, even when your affiliation changes, ensuring that you get the credit you deserve and helping you keep track of your work.�ROR, The Research Organization Registry (ROR.org) is a global, community-led registry of open persistent identifiers for research organizations. ROR makes it easy for anyone or any system to disambiguate institution names and connect research organizations to researchers and research outputs. For example, Science for Life Laboratory is commonly referred to as SciLifeLab. If not directly involved with the organisation one might easily think that these are two different organisations, using the SciLifeLab ROR-id will make it clear they refer to the same organisation.

36 of 68

Accession numbers

The European Nucleotide Archive (ENA) assigns accession numbers to digital entities at various levels of granularity. �

PIDs and accession numbers for Life Sciences data

Many databases and repositories in life sciences work with their own identifier system. As these databases are considered as trustworthy data repositories by many instances, their accession numbers are widely adopted and accepted for use in citations.
The European Nucleotide Archive (ENA) assigns accession numbers to digital entities at various levels of granularity. Example ENA accession numbers are the PRJEB1788 (BioProject) and ERR 164410(Sequence).

Other examples:

The European Genome-Phenome Archive (EGA) assigns IDs to studies and datasets. An example dataset ID is EGAD00000000001.
EMBL-EBI’s BioStudies database links to or manages diverse biological data from EMBL-EBI databases as well as data that do not fit in the structured EMBL-EBI databases. Once data are submitted successfully, an accession number is automatically assigned. Example accession numbers are S-BSST7 (BioStudies), S-BEAD712 (BioImage Archive), E-MTAB-12919 (ArrayExpress).

37 of 68

Benefits of PIDs

uniquely distinguish resources from similar objects (F)
a place to keep the metadata (F)
machine actionable identifiers increase findability (F)
resolves providing a way or information on how to access the object (A)
enhances citability leading to easier reuse (R)

38 of 68

Licenses

Open Science in the project lifecycle - access & reuse

39 of 68

Why license your research output?

Image from Flickr, licensed CC0 (https://flic.kr/p/6kHZ9r)

Open Licenses for Data, Code, and Software Pilot – CC-BY SA – 19.06.2024

40 of 68

Why license your research output?

Image from Flickr, licensed CC0 (https://flic.kr/p/6kHZ9r)

FREE

to a good home

Open Licenses for Data, Code, and Software Pilot – CC-BY SA – 19.06.2024

41 of 68

Why license your research output?

A license is a clear invitation for others to REUSE your research data or software – on your terms.

Image from Flickr, licensed CC0 (https://flic.kr/p/6kHZ9r)

FREE

to a good home

Open Licenses for Data, Code, and Software Pilot – CC-BY SA – 19.06.2024

42 of 68

Why license your research output?

Sources : 1 - 2 - 3 - 4

FAIR principles for research data 🡪 R1.1. (meta)data are released with a clear and accessible data usage license.

FAIR principles for research software 🡪 R1.1. Software is given a clear and accessible license.

Open Licenses for Data, Code, and Software Pilot – CC-BY SA – 19.06.2024

43 of 68

Why license your research output?

Make the property rights clear & access and usability rules evident

Protection of one’s right
More citations
More reuse
Greater impact

44 of 68

What type of licences have you come across?

45 of 68

Overview of available open licenses

SPDX Licenses list

Open Gouvernement license for public sector information

There are numerous licenses, some of which can be customized to meet specific needs, while others are highly specific.

Etalab license

European Union Public Licence

Eclipse Public Licence

Italian Open Data License

Open Licenses for Data, Code, and Software Pilot – CC-BY SA – 19.06.2024

46 of 68

Creative Commons

Originates from session 8 of the The ELIXIR FAIR training material by Design course. CC-BY-NC-SA

47 of 68

Creative Commons - before licensing

The licenses and CC0 mark cannot be revoked

Updates only apply for “new users”�

Only the copyrights holder can apply a CC license

If created in the job , you might not be the holder

48 of 68

Creative Commons

Core conditions:

Public domain tools:

Attribution

ShareAlike

NonCommercial

NoDerivatives

No rights reserved

PDM

No known copyright

Creative commons offer 6 licences based on 4 Core conditions that are combined into the different licences.

�The Four Core Conditions:

Attribution (BY): A figure icon (person) - You need to name the original author/creator. In all 6 licences
ShareAlike (SA): A circular arrow icon - If you adapt the work, you must share your modifications under the same or a compatible license.
NonCommercial (NC): A dollar sign with a slash through it icon - You can only use the work for non-commercial purposes.
NoDerivatives (ND): An equals sign icon - You cannot share adaptations or modifications of the work.

In addition, there are two public domain tools

CC0 (Creative Commons Zero): Often represented by the number zero in a circle. It allows creators to dedicate their work to the public domain, waiving all copyright and related rights.

Public Domain Mark: Indicates works that are already in the public domain due to expired copyright.

49 of 68

Creative commons licences

Creative commons license spectrum by Shaddim; original CC license symbols by Creative Commons. CC-BY-SA 4.0

50 of 68

Mission

coordinate and support the digitalisation of public administration.
responsible for Sweden's digital infrastructure.
to follow up and analyse the digitalisation of society.
to help the government make well-informed decisions.

DIGG is a public authority (under the Ministry of Finance) that is responsible for the digitalisation of the public administration. ��Our assignment is formulated in our instructions and in our regulation letter. We are also steered through direct government assignments, where we can be given the task of investigating an issue or developing a special solution.

The instruction contains four clear overall points that clearly summarises our mission:

We coordinate and support the digitalisation of public administration.
We are responsible for Sweden's digital infrastructure.
We follow up and analyse the digitalisation of society.
We help the government make well-informed decisions.

In the context of the open data directive from 2019, DIGG has come up with recommendations for licences for public authorities

51 of 68

Creative commons licences -recommendations

information/data not subject to copyright or IP protection

information/data subject to copyright

authorities should not impose this type of limiting condition or control compliance

The Agency for Digital Government (DIGG):

Rekommendation om öppna licenser och immaterialrätt | Digg

Creative commons license spectrum by Shaddim; original CC license symbols by Creative Commons. CC-BY-SA 4.0

The Agency for Digital Governance has come up with recommendations for �

For information and data created by a public authority that is not subject to copyright or other intellectual property protection, the marking PDM or CC0 is recommended. For databases created by a public authority for the exercise of public authority, the marking CC0 is recommended.

The recommendation presupposes that unrestricted further use can take place in accordance with other legislation.

�For other information created by a public authority that is subject to copyright protection as a work or performance, the license CC BY 4.0 is recommended. This also presupposes that the authority has acquired sufficient rights or has fully taken over the economic rights from the creator through an agreement.

The authority is advised against using the CC BY-SA license, as the condition that others' processing of such information must be licensed under the same license entails a limitation. Authorities should not impose this type of condition or control compliance.

The DIGG has more extensive recommendations also for software. Unfortunately this is only in Swedish but using a translating tool could make them useful. Please have a look at their website.

52 of 68

Tools for choosing a license

License selector�https://ufal.github.io/public-license-selector/

Choose a license�https://choosealicense.com/

Creative Commons license chooser�https://chooser-beta.creativecommons.org/

53 of 68

Licences vs Terms of use

Disadvantages:

Not standardised
Not machine readable

As we’ve seen earlier when speaking of identifiers, repositories that have been around for some time might have their own solution regarding terms of use that are not standardised licences. One example of this is EMBL-EBI data repositories.

Disadvantages:

Not standardised

Not machine actionable

https://www.science.org/doi/10.1126/science.298.5597.1333b

We will:

Standardise the licences used across EMBL-EBI resources.
Use licences that present the lowest barriers to data reuse. CC0 is preferred over CC-BY. CC0 is most in line with the spirit of the EMBL-EBI Terms of Use.
State the licence explicitly on the resource as a whole and at the record level, in both human and machine-readable formats.

In some cases EMBL-EBI’s Terms of Use may have to be retained (for example, for a data resource that aggregates content with existing multiple licence types).

54 of 68

Licences vs Terms of use

Plot from EBI https://www.ebi.ac.uk/licencing/

Monitoring progress towards standardised licensing

Standardise the licences used across EMBL-EBI resources.
Use licences that present the lowest barriers to data reuse. CC0 is preferred over CC-BY.
State the licence explicitly both on the resource and at the record level.

55 of 68

Restricted access: EGA example

https://ega-archive.org/submission/metadata/ega-schema/

Metadata record for EGAD50000000298

56 of 68

Restricted access: EGA example

https://ega-archive.org/access/data-access-committee/data-use-ontology/

Lawson et al., 2021, Cell Genomics 1, https://doi.org/10.1016/j.xgen.2021.100028

57 of 68

58 of 68

59 of 68

Lets try this out!

60 of 68

Getting Your Research Out There: A Hands-On Introduction to Open Sharing with Zenodo Sandbox

61 of 68

Reflections

Were some metadata fields difficult to provide an answer to?
Did something surprise you with the metadata requirements in Zenodo?

62 of 68

Repositories are key for share and reuse

Why using repositories?

When to start thinking about repositories?

63 of 68

Repositories are key for share and reuse

Why using repositories?

Helps you adhere to the FAIR principles

When to start thinking about repositories?

Early on

File formats
Metadata fields
Lightweight backup

64 of 68

Evaluate a repository

Things to check when evaluating:�

Are others in the community using it?
Is it easy to navigate / user-friendly?
Is there support / guidance for submission and reuse?
Is it sustainable, i.e. will the repository be around for a while?
Will the datasets obtain persistent identifiers? Is the repository itself FAIR?
What are the usage rights or license options?

65 of 68

Identify repositories

How to find a suitable repository for life science data?

EBI repository wizard - guide depending on data type
ELIXIR deposition databases - core resources with long-term data preservation and accessibility plans
FAIRsharing.org/databases - catalogue of many repositories, with possibility to filter on e.g. domain
Scientific Data Repository Guidance - publisher’s recommendation
re3data.org - registry of research data repositories (not only life science)

66 of 68

Demo: EBI Repository Wizard

Which repository would be suitable if you have a genomics project with mice RNA sequences?�

Go to https://www.ebi.ac.uk/submission/
Answer the questions regarding

data type
need for controlled access
if experimentally produced by you
type of study

67 of 68

Key Points

Rich metadata is the key to scientific reliability and reuse.
No license - no reuse (without asking for permission).
Persistent identifiers allow an object to be identified, verified and located.
Using repositories increases FAIRness and reuse:

Identifying a repository early one helps to formalise metadata from the start.
If possible, use a domain-specific repository since it has maximum reach in the research community.
The research output (data) types determines which domain-specific repository is suitable.

68 of 68

Intro DM course

nbis.se/training/introduction-to-data-management-practices