Difference between revisions of "PREMIER Reporting"

From PREMIER QMS - Wiki
Jump to navigation Jump to search
 
(No difference)

Latest revision as of 21:34, 2 December 2021

← QM House

Objectives

The aim of reporting is to make data and research results transparent, provide them to the scientific community and foster the required exchange and cooperation.

Background

Reporting of research results should preferably follow the principles of open science. Open Science combines the guiding principles of Open Access, Open Data and Open Source. Although all Open Principles can exist independently, there are synergies of the principles which can also be applied in a meaningful way in everyday scientific life.

Tasks / Actions

In order to create a lab specific action plan, the first step is an assessment, which will be carried out by the PREMIER team. The assessment will determine the status quo of the laboratory in regard to existing quality tools. Here you find the general tasks / actions that are necessary to implement the module.

Open Science

Open Access involves long-term and unrestricted access to scientific results in the form of publications / papers and scientific literature. Open Data is publicly accessible data that can be freely used, reused and shared by anyone; the only restriction is the obligation to name the author of the data. Open Source is software with publicly available source code that is freely available and subject to a recognized open source software license. Furthermore, open source means freely available knowledge and information in general. Basic research should be committed to all these principles of openness, scientific transparency and the possibility of unlimited reuse of its post-publication data. Through Open Science it is possible to guarantee full transparency of research services, including the validity of used models and the reproducibility of the data, e.g. through replication studies. This allows a time-independent, public, further evaluation, also by secondary or meta-analyses of the data far beyond a limited readership. In addition, research processes can be accelerated through better data sharing. The informal added value of each publication is increased if raw data is also available. These public raw research data are stored in suitable and recognized online repositories. Repositories are managed repositories for the storage of ordered documents that are accessible to the public or a restricted group of users. For researchers it is important to pursue a consistent publication practice. This means that all results, data and procedures of the research are made available to the scientific community and the general public as openly as possible, i.e. without access restrictions and in complete form. In all processes, it is important to observe conformity with applicable law (animal protection, occupational safety; genetic engineering, etc.), the principles of Good Scientific Practice (GSP), and commit yourselves to observing internationally recognized rules (e.g. ARRIVE Guidelines for animal experiments, and International Committee of Medical Journal Editors (ICMJE) recommendations for publications).

Open Access

Basic principles

There are several models for Open Access:

  • Golden Way

Immediate free availability in an Open Access journal after successful peer review. All work in such journals is Open Access, e.g. journals of the PLoS series. Incomplete List of open-access journals:List of open-access journals

  • Green Way

Secures the author the right to parallel publication or self-archiving of an original work, in addition to publication by the publisher. The author grants the publisher the simple right of publication, but there is no transfer of all exploitation rights (no copyright transfer to the publisher). According to the publisher's regulations, the author may make either the content of the article or the PDF of the article available to the public via:

  1. institutional repositories (e.g. document server of a university (for the Charité:refubium.fu-berlin.de/at the FU Berlin)
  2. scientific repositories (servers of a thematic orientation)
  3. own homepage (recommended with a Creative Commons license)

A list of scientific Open Access repositories can be found at OpenDOAR and in the Registry of Open Access Repositories (ROAR).

  • Postprint

Many publishers or journals have conditions according to which the content of a peer-reviewed publication may be archived by the authors themselves. The conditions for this are sometimes more relaxed (e.g. the publisher's PDF may be used) or sometimes more restrictive (only the accepted manuscript may be published). Sometimes embargo times have to be taken into account. The exact conditions under which accepted manuscripts may be published can be found in theSHERPA / RoMEOdirectory at the entry of the respective journal.

  • Hybrid Model

The non-Open Access publisher grants Open Access rights to self-archive the publisher's version upon payment of a fee by the submitting author. Hybrid is this model because OA is the exception rather than the rule and an additional source of income in such journals, in addition to the parallel subscription. The publisher thus collects twice: once from the author, who wants to make his publication accessible to all, and once from the institution, the fee for institutional access to all articles. This model is supported by fewer and fewer founders.

  • Green Way through Funders Mandate

Some non-Open Access journals grant Open Access rights to original works funded by certain large public financiers. BMBF and DFG are not among them.

PREMIER recommends both, the golden and the green path.The decision as to which path to take lies with the publishing author or the working group.

  • Bronze Way

Journals offers delayed open access, i.e. after an embargo period.

  • Diamond / Platinum Way

Journals do not charge either readers or authors but require funding from external sources.

Publishing Open Access

Even before writing the manuscript of the original work, the corresponding author should choose a journal that allows Open Access publishing. For journals that support the golden path, the QUEST Center of the BIH has developed a curatedOpen Access Journal Whitelist. It contains journals with high quality standards in order to protect against publications in so-called predatory journals. In addition, the list, which can be sorted by subject area, contains information on the review times, publication costs and whether the DFG Publication.

Alternative sources of information for Open Access journal selection are:Sherpa/RomeoandDirectory of Open Access Journals.

After acceptance of the original work, the author must apply to the publisher for Open Access publication via the green or hybrid way. The journal usually sends the necessary forms to the corresponding author, otherwise this must be requested.

  • „Predatory“ / Junk Journals

Open Access has come into disrepute with the emergence of so-called "predatory" or "junk" journals. These Predatory Publishers often invite you via e-mail to submit manuscripts for which they charge Open Access fees, which at first glance seem relatively low. However, these manuscripts often do not undergo an orderly and recognized peer review process, nor do the publications appear later in PubMed, making them less visible and often qualitatively worthless.

  • Higher-level open access specifications

TheBerliner Senatcalls for at least 60% Open Access from all public scientific institutions in 2020. The BMBF is primarily committed to the green path and sets Open Access as the standard in its own funding programs. The DFG also recommends and supports efforts towards Open Access publications. For EU-funded Horizon 2020 projects, Open Access is mandatory for all publications (including books and monographs). An overview of the Open Access policies of various funding organizations can be found in theSHERPA/Juliet Database

  • Subsequent secondary publication for existing publications (Postprint)

There is a legally secure way to make the contents of publications subsequently accessible to the public. First, you should check whether the article has free full text access after all. (e.g. via Text availability --> "free full text" in theSide bar filter in PubMed)) or in thePMCcatalogue, which only contains openly accessible articles. Uploading a PDF article, for which the publisher has not expressly permitted self-publication, to public social platforms such asResearch GateorAcademia.euconstitutes a copyright violation, which may result in criminal prosecution for the uploader. However, sending a private copy by e-mail is permitted. The German secondary publication law(deutsche Zweitveröffentlichungsrecht)also permits the publication of the accepted manuscript version 12 months after the first publication on institutional repositories for projects that are funded mostly with public resources. Documents posted there are indexed byGoogle Scholarand Open Access search engines such asbase-search.netandunpaywalland thus, permanently searchable. Alternatively, the EU's ownZenodorepository offers the possibility of publishing on the green path of Open Access. TheSherpa-RoMEODatabase provides journal-specific information on which publication form (Postprint PDF, Publisher PDF or Preprint PDF) may be used for the repository upload after the embargo period.

  • Pre-publication of manuscripts (Preprints)

It often takes months after the submission of a finished manuscript for publication, sometimes even years for rejection and new submission. This is remedied by a new means of communication that is becoming more and more popular: The purpose of preprints is to make publication content public immediately, e.g. before or during the submission of the manuscript for peer review to a journal. More and more publishers and journals are accepting this approach (List of Academic Journals by Preprint policy). Preprints are uploaded to special preprint servers. The following preprint servers are recommended for biomedical manuscripts:

  1. bioRxiv
  2. medRxiv
  3. OSF

Preprints offer many advantages:

  • The manuscripts can be uploaded to the preprint servers free of charge.
  • The published content is up to date. Especially with scientifically hot topics, a preprint can give you a head start in the community.
  • Since all preprints receive aDigital Object Identifier (DOI), preprints can also be permanently quoted and found.
  • Many publishing houses (e.g. Elsevier, Springer-Nature, PLoS) allow cited preprints in manuscripts submitted to them to be officially listed as references in the bibliography.
  • Many traditional journals allow that manuscripts submitted to them for peer review are already available as preprints.(List of journals that accept preprints).
  • Even if the resulting peer-reviewed publication is not Open Access, the content remains accessible worldwide via the preprint.
  • Important funding organizations such asNIH,Wellcome und MRCrecognize preprints in funding applications as citable works.

Disadvantages of preprints:

  • Contents are not peer reviewed; the significance of preprints is therefore limited.
  • not listed in PubMed

Qualitatively poor works find a way to reach scientifically oriented audiences.

Open Data

Basic principles

Open Data in scientific research means reliable access to the underlying original data for each interested person and in the sense of maximum transparency, quality assurance of possible reproducibility and cost savings through possible secondary analyzes. The provision of raw data for subsequent use is to be carried out promptly and completely. Anyone may use, share or modify the disclosed data for any purpose.

A growing number of journals demand transparency in the publication and the disclosure of the raw data, and increasingly, sponsors demand this as well. It is to be expected that from 2020 onwards, research grants will no longer be awarded unless a data management plan is included. The DFG now requires"making research data available as soon as possible," immediately after completion of the research or after a few months… "Research data should be accessible at an early stage, allowing meaningful and further use by third parties.

What should be observed before the data are published?

  1. Who created or commissioned the creation of the data?
  2. Is there any personal data in the data set? If so, anonymization must be ensured, in accordance with the Berlin Data Protection Act.
  3. Is the data usable for others? Exact captions and descriptions in separate files are essential.
  4. Are the data only available in proprietary formats? In this case, a file copy (for example, by export) should also be made public into a generally accessible open format.

How does Open Data work?Each staff member is free to use a repository of choice for Open Data. An overview of existing repositories (> 1700) can be found in the DFG-funded projectre3data.org.For general, multidisciplinary data, Experimental Neurology recommends three international online repositories (Zenodo, Figshare, Mendeley Data), in which:

  • Raw data (data of all types, tables, diagrams, pictures, but also posters) can be permanently identified and citable by a permanent Internet address DOI (Digital Object Identifier).
  • The data volumes for freely available data are unlimited.
  • An embargo can be placed on data deposited before it becomes visible to the public.
  • It is possible to link the uploader with ORCID.
  Zenodo.org Figshare.com data.mendeley.com
maximum file size 50 GB/dataset 5 GB ?
Upload size private 50 GB 20 GB ?
embargo for public data yes, selfset limit up to 12 months 6, 12 month or selfset limit
data findable in search engines https://share.osf.io;google.com;opendoar.org;www.base-search.net https://share.osf.io;google.com;opendoar.org;www.base-search.net google.com;www.base-search.net
re3data.org registry www.re3data.org/
repository/r3d100010468
www.re3data.org/
repository/r3d100010066
www.re3data.org/
repository/r3d100011868
Data upload no instructions https://support.figshare.com/
support/solutions/articles/
6000102511-how-to-upload-your-data
no instructions
Data publishing no instructions https://support.figshare.com/
support/solutions/articles/
6000091698-how-to-publish-your-data
no instructions

To use one of these three services, a previous registration is necessary.

Special Data

These include, for example, -omics, open fMRI, RNA / DNA and protein sequence data. These are stored in special databases, which have their own requirements for data formats and metadata. A selection of suitable databases can be found at:Fairsharing

Personal data

In the event that Open Data with personal data is considered, the document "Basic information on Open Data publications in studies with personal data" provides an overview of the steps to be taken if the free availability of personal data is planned. This requires, among other things, a positive vote by the ethics committee and a correspondingly effective consent. In addition, some basic concepts such as de facto anonymization are explained, exemplary standard formulations for patient consent are given, and alternatives are described for cases where the complete disclosure of data is not possible.Some personal data (records) cannot be made openly available without further explanation due to legal or contractual provisions, for data or other protection reasons or for ethical reasons. However, it should be checked whether it is possible to make this data at least accessible in a limited way, i.e. non-critical, anonymizable data that is not subject to any obligation of secrecy.

Open Data Principles

TheFAIRprinciples define minimum standards of data to make them easier to find, accessible, fully compatible, and reusable (Findable,Accessible,Interoperable,Reusable). The following section describes how this requirement can be implemented.Open Data: General ApproachSimple, undocumented upload of the original data to one of the repositories is a procedure no longer recommended. One of the main objectives of Open Data is that data in repositories can be comprehensible and human as well as machine-readable. Usually, only one container file (Zip / Archive) is stored in the repository. This contains all necessary files, including an explanatory metafile. A unique DOI is created and the work is thus clearly identified.Proprietary formatsAny proprietary formats, such as raw image files or output files from devices that are readable only with licensed and non-open accessible software, should be avoided or converted into open, license-free formats by means of converted copies. For proprietary formats, the specification of the software, version, manufacturer and operating system is essential. Examples of non-proprietary open formats are:

  • CSV - comma separated value (tables)
  • XML - Extensible Markup Language (documents, texts with a hierarchical structure
  • Gzip / ZIP compression file format
DO DO NOT
Exact and concrete title Merge cells
Each column should have a heading and be limited to one cell Use special characters or spaces
The first cell of a data sheet (A1) must not be free Use formatting or color-coding
The file name should reflect the content as precisely as possible Merge cells
Each file may contain only one data sheet  
In each data sheet only one table  
Each cell should be occupied, missing values are marked with NA (not available) and undefined values (e.g. division by zero) with NaN (not a number)  
All acronyms used, units are transferred to the metafile and defined there  
General table rule: Variables as columns, samples / observation as rows  
Use of unique identifiers for samples / observation  
Exact and concrete title  

Images

Image files of digital cameras or microscopes are often stored as raw image files. Although these are raw data, they are usually proprietary software (see raw filename extensions and the camera manufacturers). The inclusion of information about, the manufacturer / software / version / operating system is mandatory when storing these files. Including an additional export in TIFF format, which makes minimal compression changes, is an even better complement. Associated image sequences are stored in folders, named in a way that describes the included files accurately. Further descriptions (e.g. origin, reference, etc.) of the files / folders should be included in the metafile.

Source Code

GitHub is the best-suited Internet hosting service for Open Data. A GitHub link with another repository (e.g. Zenodo) is easy to set up, as it is often useful to store the records and the corresponding code together. The following procedure is recommended to make quotable a public GitHub repository via Zenodo:

  1. Log in to Zenodo using the GitHub access data
  2. Set-up theZenodo-GitHubSynchronization. Zenodo can only archive public repositories.
  3. A new Zenodo entry is automatically triggered after the release of the GitHub Repository.
  4. Take the DOI badge from Zenodo and add it to the readme file of the GitHub repository.

Metadata

Metadata is collected text information about a data set that describes its specific characteristics in a sufficiently comprehensible manner, so that independent reuse is ensured. Parts of these can also be used in the respective data description fields of the repositories. We recommend a minimum standard consisting of:

  • Title: a descriptive title of the data and all associated work in the file store
  • Author: Name (s) of the data author
  • Creation date: as dd/mm/yyyy
  • Method: a description of how the data was created (for example, devices that provided the data used experimental method)
  • Description: A description of each data record (including description of the variables and their units, acronyms, etc.)
  • Discipline, Keywords, Reference Quotes
  • Validation: All necessary information, using tools and instruments that validate the published results (e.g. formulas, valid measuring ranges, analysis protocols, algorithms, software code, etc.)

Open licenses

In order to ensure a subsequent use of the data, the data must have an open license. This is usually done by the repository, but sometimes it is asked. Recommended open licenses are CC0 or CCBY.

Citation and link of records in articles

A data set that stored in one of the recommended repositories is given a permanent DOI and can be quoted as such. The DOI of the data set should not appear in the text of a publication or as a supplement, but rather be referenced in the bibliography. The following information is mandatory: authors, publication date, title, publisher of the record, to the record. The generated reference could then look as follows:First Author, Second Author ... Last Author (2016) Data from "Dataset title". Zenodo repository. The prefix: https://doi.org/ ensures that the journal or the publication recognizes this address as an active link. During the review process, reviewers can get confidential access to the stored data, even before the data is published. When publishing or accepting the manuscript, the data must then be openly disclosed, and the restriction actively canceled. If the final URL of the article is determined, this should be included in the expanded properties of the record.Data Management Plan Research sponsors such as Horizon 2020 of the EU, are increasingly calling for data management plans (DMP). As a helpful reference are the documents of the Charité Research Group. Moreover, there is a tool for creating DMP. The DMP Tool is a web-based tool that helps you construct data management plans using templates that address specific funder requirements.

References