🔗 This project is part of a larger open data ecosystem. Find out more information here!
🌍 Learn more about this project on our website
Uniform and FAIR data management for scientists:
📤 Remove the hassle of uploading meaningful research data
📓 Safeguard rules to ensure uniformity within and between organizations
♻️ Streamline storage and access of research data
Data management assistant for Biodiversity Observation Networks (DMBON-assistant) is a case study of a more general data management framework and webtop application that supports scientists by providing a consistent and streamlined dataflow within and between individuals and organizations. It lowers the barrier for collaborative efforts by making data more accessible and searchable, enabling more complex search queries on open datasets that enrich meta-analyses. The DMBON-assistant is an important piece of our grander vision.
This project is intended for academics, policy makers, and the general public. Academics are encouraged to share their research in a way that makes their research more meaningful and easily citable. Policy makers benefit from this tool through lowering the barrier for extensive meta-analysis that can shed light on previously unreported correlations. Finally, the general public benefits from such an open source tool by making academic research more easily accesible to the public. We believe that open sourcing a project such as this benefits the community and with it we hope to advocate for future open science policies to come.
In this README, we will explain our vision for this project. We will further motivate the choices that we have made and further provide both users and developers on how to use and contribute to DMBON-assistant.
| ➡️ | Different README files are located within this project, as github renders files with such a name nicely below the file information. Please do not confuse other README files in subdirectories with this one. |
|---|
We believe that discovery relies on open and transparent management of data by following the FAIR principles.
We furthermore believe that researchers should be able to search through scientific data just as easy as if they are performing a web-search.
This means that search queries such as "what was the water temperature during recent sightings of killer whales in the North Pacific Ocean?", should be made possible.
Obviously, for scientific data to be FAIR and meaningful to allow such queries, it has to adhere to specific rules such that others can discover, filter and (re-)use the data with little effort.
Ideally, tools should exist such that those that gather data should be able to upload and distribute meaningful data with relative ease.
This project aims to do exactly that: assist people in uploading their data (including metadata) to the web.
We believe that future projects, potentially built on top of this one, can then further support complex search queries on distributed datasets.
✔️ Introduce a simple assisted procedure for uploading heterogeneous data that conforms to FAIR data management plans.
✔️ Enable complex search queries for meta-analysis on distributed, heterogeneous databases.
✔️ Build a thriving community that supports open science.
Sound scientific results are derived from good data. While few people disagree with this statement, the current methods of both storing and accessing historical data are severely hampering novel scientific discoveries. On one end, data of published studies are permanently lost (Vines et al., 2014) or researchers forget the details requires to make use of the data (Whitlock et al. 2010). On the other end, losing access to data (and, importantly, their metadata) makes it impossible to reproduce scientific results or perform critical meta-analyses. Therefore, automated streamlining of data management plans (DMPs) is necessary for improving data preservation and accesibility. DMBON-assistant is an open source webtop application designed to provide this automisation, thereby enabling consistent and streamlined data management.
One critical, but often overlooked, feature that greatly increases the longevity of datasets is metadata (Leipzig et al. 2021), i.e. 'data that provides information about other data'. Metadata often encodes how to interpret the actual data, e.g. descriptions of meaning of columns in a table, and is thus often necessary for meta-analyses. While sharing primary data is becoming increasingly more common (even though the process is not streamlined, see, e.g. Alsheikh-Ali et al., 2011 and Baker, 2015), extensive efforts are often needed to gain access to metadata even when data is openly available (Tedersoo et al., 2021). DMBON-assistant aims to alleviate this issue by assisting its users in providing the sufficient metadata that is needed to correctly interpret the data.
We are a small team and can only do so much.
Furthermore, we also cannot discover and fix all bugs and we can never cover all potential use cases and develop features for them.
We are additionally huge advocates for open science and thus open source projects, such as this one.
Therefore, we lay our hopes and trust with you, the community, to improve and further extend DMBON-assistant's functionality.
We welcome any and all suggestions, feedback, bug reports and general discussion on DMBON-assistant.
If you are interested, please find the section on contributing below!
Interested? Great! But how does DMBON-assistant actually help those that gather data to upload and uplift their data in a meaningful way? In general, the assistant (as the name suggests) aims to assist in the data upload process to ensure uniformity both within and between organizations. Our general envisioned flow is illustrated below:
On the left, we see a user that has multiple sources of data and metadata. These can range from raw data and metadata to (source code to) analyses and publications. Importantly, much of the raw data is often uploaded to some external cloud service (such as ENA for DNA sequence information). As such, links to the raw data can be provided as well!
DMBON assistant, here highlighted in blue, provides an easy way of uploading heterogeneous data using any browser! Hidden under the hood, DMBON assistant ensures that the data description conforms to the data management plan of the organization. Finally, the assistant generates a publishable Research Object Crate (RO-crate) that neatly packages heterogeneous research data with their metadata. Explain the general premise of DMBON-assistant
As such, using DMBON assistant greatly simplifies generation of FAIR research object crates by:
- providing an easy to use interface to upload research data using your browser
- ensuring correspondence with data management plans
Frontend
For general use we do not recommend building from source, but for advanced use and/or adaptation (i.e. as a contributor) it can be useful to build from source instead.
The frontend is written in JavaScript using RedwoodJS.
We recommend installing all necessary packages/libraries with yarn as the package manager for Node.js.
- Install Node Version Manager to ensure the corrent version of Node will be used to support building the frontend:
$ wget -qO- https://raw.githubusercontent.com/nvm-sh/nvm/v0.39.1/install.sh | bash- Install Node.js (>=14.19.x <=16.x)
$ nvm install 16- To manage the frontend, use Yarn which, in turn, is managed through Corepack. Versions of Node.js <16.10 do not automatically include Corepack, so run:
$ npm i -g corepackIf the version of Node.js >=16.10, Corepack is included by default but needs to be enabled:
$ corepack enableThen, activate Yarn:
$ corepack prepare yarn@stable --activate- Run installation scripts:
How to quickly get started using DMBon-assistant, using a very simple usecase.
Explain the general workflow that users should adhere to
Highlight best practices (i.e. DOs & DONTs). Do not be too restrictive!
Please find our CONTRIBUTION guidelines for more details on the process of contributing.