RDM_IntroRDM.qmd

---
title: "Introduction to Research Data Management (RMD)"
subtitle: "A primer for researchers"

author: 
  - name: Daniel Manrique-Castano, Ph.D
    email: daniel.manrique-castano@alliancecan.ca
    affiliations:
      - name: Digital Research Alliance of Canada

date: last-modified
date-format: full

format: 
  html:
    embed-resources: true
    mermaid:
      theme: forest
      
  revealjs:
    footer: "Introduction to Research Data Management (RMD) - Daniel Manrique-Castano, Ph.D"
    logo: "images/alliance_logo.png"
    theme: white
    transition: fade
    slide-number: true
    show-slide-number: all
    preview-links: auto
    mermaid:
      theme: forest
      mermaid-format: svg
    
filters:
  - reveal-auto-agenda
auto-agenda:
  bullets: numbered
  heading: Agenda
  
  gfm:
    mermaid-format: svg

engine: knitr

css: styles.css
bibliography: references.bib 
editor: source
---

# Contextualizaing Research Data Management (RDM)

## Why “bother” with RDM?{.center}

[Data]{style="color:green;"} is the [foundation of scientific research]{style="text-decoration: underline;"}. We want to best manage this source for modeling, workflow automation, quality control, and data sharing with our field to produce meaningful scientific work. 

::: callout-tip
## The 80/20 rule

Researchers spend [80% ]{style="color:green;"} of their time finding, producing, cleaning, and organizing data.
:::

## We want our research de be reproducile{.center} 

::: {style="text-align: center; font-size: 70%"}
![Reports on the reproducibility crisis](images/crisis1.png){fig-align="center" width="1200" height="500"}
:::

## Some well-known causes

::: {style="text-align: center; font-size: 70%"}
![Causes of the reproducibility crisis](images/crisis.png){fig-align="center"}
:::

## Creating reproducible science{.center} 

{{< bi file-earmark-text >}} [Document data provenance]{style="color:green;"} (samples, software, instruments, other tools).

{{< bi gear >}} Define [standard procedures and methods]{style="color:green;"} (i.e., how measurements were made).

{{< bi funnel >}} [Quality control]{style="color:green;"} measures (data evaluation, organization and cleaning).

{{< bi file-earmark-text >}} [Comprehensive documentation]{style="color:green;"} to enable understanding and reuse of data.


## Therefore, research needs to move towards{.center} 

- [Competent]{style="color:green;"} researchers in [RDM]{style="text-decoration: underline;"} and data analysis.

- [Standardized approaches]{style="color:green;"} to [sharing]{style="text-decoration: underline;"} raw data and analysis code to support research findings.

- Researchers with a [commitment to transparency]{style="color:green;"} and best scientific practice practices to ensure [research integrity]{style="text-decoration: underline;"}.

::: callout-tip
## RDM competences

Think of RDM competencies as a matter of professional responsibility, ethical conduct, and commitment to best scientific practices.
:::


## Tri-Agency Research Data Management Policy{.center} 

The [Goverment of Canada](https://science.gc.ca/site/science/en/interagency-research-funding/policies-and-guidelines/research-data-management) promotes RDM in its [Tri-Agency Research Data Management Policy](https://science.gc.ca/site/science/en/interagency-research-funding/policies-and-guidelines/research-data-management/tri-agency-research-data-management-policy-frequently-asked-questions).

Through its federal funding agencies, the  the Government of Canada seeks to implement [data management plans (DMPs)]{style="text-decoration: underline;"} and [sharing]{style="text-decoration: underline;"} of research data to maximize the benefits to society. 


## We can define RDM as: {.center}

[RDM]{style="color:green;"} encompasses the organization/handling, storage, sharing and preservation of research data, including physical objects/files, documents, images, audio, digital databases, etc.


# Understanding RDM

## Where does data end up?{.center}

::::: {layout-ncol="4"}
:::: {#first-column}
::: {style="text-align: center;font-size: 60%"}
![Laptop of students and postdocs](images/Laptop.jpg){fig-align="center" width="100%"}
:::
::::

:::: {#second-column}
::: {style="text-align: center;font-size: 60%"}
![Institute network](images/Network.jpg){fig-align="center" width="100%"}
:::
::::

:::: {#third-column}
::: {style="text-align: center;font-size: 60%"}
![The cloud (Google drive)](images/Cloud.jpg){fig-align="center" width="100%"}
:::
::::

:::: {#forth-column}
::: {style="text-align: center;font-size: 60%"}

![HPC cluster](images/HighCompuating.jpg){fig-align="center" width="100%"}
:::
::::
:::::


## Considerations for storing data{.center} 

{{< bi clock >}} In most cases, data must be retained for [10 years]{style="color:green;"}.

{{< bi safe >}} What do we need to [store]{style="color:green;"}?

{{< bi file-earmark-text >}} What [data formats]{style="color:green;"} provide access?

{{< bi gear >}} What [technical requirements]{style="color:green;"} do I need?

{{< bi unlock >}} How [secure]{style="color:green;"} is the data and how can I access it?

::: callout-tip
These are all aspects of good scientific practice!
:::


## What research data we need/could store?{.center} 

{{< bi thermometer-snow >}} [Experimental samples]{style="color:green;"} (physical samples and files, and the methods for generating them).

{{< bi robot >}} [Scientific instruments]{style="color:green;"} (model, version, setup).

{{< bi journals >}} [Experimental data]{style="color:green;"} (raw and processed data, plots, tables and figures).

{{< bi code >}} [Code and analysis pipelines]{style="color:green;"} (processing and analysis workflows, software and libraries, machine learning classifiers, statistical models).


## {.center} 

::: {style="text-align: center; font-size: 70%"}
![Image by https://biosistemika.com/](images/FAIR_Principles.png){fig-align="center" width="700" height="350"}
:::

## {.center} 

::: {style="text-align: left;font-size: 80%"}
{{< bi search>}} [**Findable:**]{style="color:green;"} Data with [identifiers and metadata]{style="text-decoration: underline;"} that facilitate its discovery.

{{< bi unlock>}} [**Accessible:**]{style="color:orange;"} To the community and research under clear [permissions]{style="text-decoration: underline;"} (i.e [Creative Commons Licenses](https://creativecommons.org/share-your-work/cclicenses/)).

{{< bi upc-scan>}} [**Interoperable:**]{style="color:blue;"} Data  conforming to [specific standards]{style="text-decoration: underline;"} (i.e. standardized vocabularies) and [formats]{style="text-decoration: underline;"} (i.e. CSV, TIFF).

{{< bi recycle>}} [**Reusable:**]{style="color:magenta;"} Data [sufficiently described]{style="text-decoration: underline;"} (and licensed) to allow reuse by interested parties.
:::

::: callout-tip
Additional information and guidelines can be found [here](https://www.go-fair.org/fair-principles/)
:::

# Research project life cycle

## Research project life cycle{.center} 

::: {style="text-align: center; font-size: 70%"}
![Image by SZF TU Berlin](images/DataCycle.webp){fig-align="center" width="600" height="400"}
:::


## Planning phase{.center}

:::: {style="text-align: left;font-size: 70%"}
Planning includes [clear perspectives]{style="color:green;"} on how researchers will [manage, store, and share]{style="text-decoration: underline;"} their data. This step may include:  

{{< bi clipboard-data>}} **Identify data needs:** Determine what [type of data]{style="text-decoration: underline;"} is needed or available to answer the research question.  

{{< bi file-earmark-text>}} **Create a Data Management Plan (DMP):** Outline how data will be [collected, documented, and preserved]{style="text-decoration: underline;"}, file formats, appropriate [metadata standards]{style="text-decoration: underline;"}, and long-term data [sharing and preservation]{style="text-decoration: underline;"}. 

::: callout-tip
## DMP assistants

The [DMP Assistant](https://alliancecan.ca/en/services/research-data-management/dmp-assistant) from the [Digital Research Alliance of Canada]{style="color:orange;"} is an example of a DMP tool. 
:::

{{< bi file-person>}} **Ethical considerations:** Address any ethical requirements related to the data [privacy, consent, and security]{style="text-decoration: underline;"}.
::::

## Collection phase{.center}

:::: {style="text-align: left;font-size: 80%"}
Researchers focus on [collecting research data]{style="text-decoration: underline;"} from:

- **Experiments:** laboratory procedures or clinical trials.

- **Fieldwork:** population surveys or environmental measurements.

- **Existing datasets:** secondary data from repositories or historical archives). 

::: callout-tip
## Consider:

- Standards for data collection.
- Technical elements (what data is stored and how).
- Adequate documentation and relevant metadata.
:::
::::


## Storing phase {.center}

:::: {style="text-align: left;font-size: 80%"}
Properly store research data to prevent loss, corruption, or unauthorized access. Some [best practices]{style="color:green;"} include:

{{< bi database>}} **Systematic backups:** Having [multiple copies]{style="color:green;"} of the data in multiple physical and digital locations.

{{< bi file-lock2>}} **Security:** Sensitive data must be [protected]{style="color:green;"} from unauthorized access by encrypting files or restricting access to authorized personnel.

{{< bi bar-chart-steps>}} **File organization:** Appropriate and consistent [naming conventions]{style="color:green;"} to facilitate the location and understanding of records.
::::

::: callout-tip
## Active RDM platforms

Consider using active data management platforms suah as the [Open Science Framework](https://osf.io/) (OSF) or [OpenBis](https://openbis.ch/).
:::


## {.center}

::::: {layout-ncol="2" layout-valign="center"}
:::: {#first-column}
:::{style="text-align: left;font-size: 80%"}
For example, [TIER 4.0](https://www.projecttier.org/tier-protocol/protocol-4-0/root/) is [systemic template]{style="color:green;"} to standardize an appropriate folder structure for research data. 

You can download the template [here](https://github.com/Alliance-RDM-GDR/RDM_IntroRDM/tree/main/resources).
:::
::::

:::: {#second-column}
::: {style="text-align: center;font-size: 50%"}
![Folder tree](images/tier.png){width="35%"}
:::
::::
:::::

## Analyze phase{.center}

:::: {style="text-align: left;font-size: 80%"}
Analyzing data to draw scientific conclusions. This may include:

{{< bi clipboard-data>}} **Data cleaning:** Using [reproducible workflows]{style="color:green;"} to handle and clean data, removing possible errors, duplicates, and inconsistencies.

{{< bi code-square>}} **Statistical analysis:** Use theoretically motivated [statistical models]{style="color:green;"} to make scientific inference and identify patterns or relationships in the data. 

{{< bi image>}} **Visualization:** Create visualizations to [present scientific results]{style="color:green;"}.

::::

::: callout-tip
## Analysis notebooks

Approximations using open source [Jupyter](https://jupyter.org/) (Python) or [Quarto](https://quarto.org/) (R and Python) notebooks are recommended for replication purposes.
:::

## Partners to handle analysis workflows{.center}

::::: {layout-ncol="2" layout-valign="center"}
:::: {#first-column}

### {{< bi code-square>}} R-studio/Quarto (R + Python)
::: {style="text-align: center;font-size: 100%"} 
![R-studio/quarto screen](images/R-studio_Screen.jpg){fig-align="center" width="500" height="300"}
:::
::::

:::: {#second-column}
### {{< bi github>}} GitHub (Version control)
::: {style="text-align: center;font-size: 100%"} 
![GitHub screen](images/GitHub_screen.jpg){fig-align="center" width="500" height="300"}
:::
::::
:::::


## With R-studio (R and Python) you can {.center}

::::: {layout-ncol="2" layout-valign="center"}
:::: {#first-column}

### {{< bi code-square>}} R-studio/Quarto (R + Python)
::: {style="text-align: center;font-size: 100%"} 
![R-studio/quarto screen](images/R-studio_Screen.jpg){fig-align="center" width="500" height="300"}
:::
::::

:::: {#second-column}
:::{style="text-align: left;font-size: 75%"}
{{< bi file-earmark-spreadsheet-fill>}} Handle [data tables and variables](https://github.com/elalilab/Stroke_PDGFR-B_Reactivity/blob/main/Widefield_10x_ROIs_CD31-Pdgfrb_Coloc.qmd) using the R [Tidyverse](https://www.tidyverse.org/).  

{{< bi images>}} Analyze [images](https://github.com/elalilab/Stroke_PDGFR-B_Reactivity/blob/main/Confocal_40x_ROIs_Pdgfrb_Morpho_BatchScript.qmd) using Python [skimage](https://scikit-image.org/). 

{{< bi wallet-fill>}} Process [Flow cytometry](https://github.com/elalilab/Stroke_PDGFR-B_Reactivity/blob/main/FlowCytometry_Pdgfrb_Processing.qmd) files/data using R [FlowCore](https://bioconductor.org/packages/release/bioc/html/flowCore.html) from [BioConductor](https://bioconductor.org/).

{{< bi bandaid>}} Analyze [RNS-seq](https://www.bioconductor.org/packages/release/bioc/vignettes/DESeq2/inst/doc/DESeq2.html) data using R [DESeq2](https://bioconductor.org/packages/release/bioc/html/DESeq2.html) from [BioConductor](https://bioconductor.org/).

{{< bi bar-chart-line-fill>}} Do state-of-the-art [statistical modeling](https://github.com/elalilab/Stroke_PDGFR-B_Reactivity/blob/main/Widefield_5x_Ipsilateral_Gfap_IntDen.qmd) using [brms](https://paulbuerkner.com/brms/).

And anything else you can think of...
:::
::::
:::::


## Description phase{.center} 

::: {style="text-align: left;font-size: 80%"}
Document the [content of the dataset]{style="text-decoration: underline;"} to make it easier to [understand and reuse]{style="text-decoration: underline;"}. This comprises the generation of a readme file containing [Metadata]{style="color:green;"} schemes (information about the data).
:::

::::: {layout-ncol="2" layout-valign="center"}
:::: {#first-column}
::: {style="text-align: center; font-size: 50%"}
![From https://github.com/twbs/bootstrap-rubygem](images/readme.webp){fig-align="center" width="700" height="300"}
:::
::::

:::: {#second-column}
::: {style="font-size: 70%;"}
There are templates/resources to guide the generation of readme files:

- [Creating a README file](https://blog.datadryad.org/2023/10/18/for-authors-creating-a-readme-for-rapid-data-publication/)

- FRDR readme file [template](https://www.frdr-dfdr.ca/docs/txt/README.txt).

- [Readme.so](https://readme.so/)

- [Readme.ai](https://readme-ai.streamlit.app/)
:::
::::
:::::


## Contents of a readme file{.center} 

:::{.r-fit-text}
In general, a dataset readme file shows:

{{< bi database-fill >}} [Dataset identifier]{style="color:green;"} showing information such as title, authors, data collection date, geographic information.

{{< bi folder-fill >}} A [map of files/folders]{style="color:green;"} defining the contents and hierarchy of folders and subfolders, along with naming conventions.

{{< bi gear-fill >}} [Methodological information]{style="color:green;"} presenting methods for data collection/generation, analysis, and experimental conditions.

{{< bi sd-card-fill >}} A set of [instructions and software]{style="color:green;"} for opening, handling and reproducing research and analysis workflows.

{{< bi credit-card-2-front-fill >}} [Sharing and access information]{style="color:green;"} detailing permissions and terms of use.
:::

::: {.callout-caution collapse="true"}
A dataset is a [standalone object]{style="text-decoration: underline;"}. Methodological information [MUST NOT]{style="color:red;"} be relegated to associated research articles.
:::


## Archiving and Sharing Phase{.center} 

::: {style="text-align: left;font-size: 90%"}
Archiving and sharing research data involves its long-term preservation in trusted data repositories, ensuring that the data are:

{{< bi search>}} **Accessible:** Data is [findable and accessible]{style="color:green;"} for reuse.

{{< bi database>}} **Preserved:** To [prevent data loss]{style="color:green;"} due to software or hardware obsolescence. Repositories have systematic process to ensure that data is always accessible.

{{< bi safe2>}} **Licensed:** to define how others can [use the data]{style="color:green;"}.

{{< bi link-45deg>}} **Cited:** with a [DOI]{style="color:green;"} or other associated unique identifier.
:::


## We can do better science{.center} 

::::: {layout-ncol="2" layout-valign="center"}
:::: {#first-column}
::: {style="text-align: left; font-size: 70%"}
**Data availability statement**

"The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation."
:::
::::

:::: {#second-column}
::: {style="text-align: center; font-size: 70%"}
![Researchers do not share data](images/NoData.png){fig-align="center" width="600" height="100"}

::::
:::::
::::::

::: callout-tip
We need to be more aware of our responsibilities and duties as scientists!
:::


## Sharing data (in repositories){.center}

::: {style="text-align: left;font-size: 80%"}
When you share data, [make sure]{style="color:green;"} that it meets these characteristics:

{{< bi diagram-2-fill >}} [Folders and files]{style="color:green;"} are organized in a [structured way]{style="text-decoration: underline;"}: Use [standardized file formats]{style="color:green;"} (e.g., CSV, TIFF) and check for consistency in [naming conventions]{style="color:green;"}.

{{< bi file-earmark-text-fill >}} [The metadata/readme]{style="color:green;"} allow the dataset to be understood as a standalone object, providing [collection methods, processing steps, and relevant context]{style="text-decoration: underline;"}.

{{< bi code-square >}} It is [desirable]{style="color:green;"} that it contains reproducible workflows used to process and generate the [research results]{style="text-decoration: underline;"}.
:::

## {.center}

Be aware that datasets are research elements that [serve the public and the scientific community]{style="color:green;"}, and that can be used (and cited) [independently]{style="color:green;"} of a research article.

::: callout-tip
## Why not?  
Think of research articles as additions to your dataset!
:::

## Reuse

::: {style="text-align: left;font-size: 70%"}
Researchers and the public can reuse data to:

-   **Avoid unnecessary or costly experiments** by using previous research results.

-   **Validate research findings:** [Independent verification]{style="color:green;"} of scientific results and conclusions (by replicating research workflows).

-   **Repurpose data:** Use the data for [new research questions]{style="color:green;"} or in combination with other datasets. They are also extremely valuable as [educational resources]{style="color:green;"}.

-   **Build upon previous work:** to [accelerate scientific discovery]{style="color:green;"} and meta-analysis by avoiding duplication of efforts or reliance on [irreproducible research]{style="color:red;"}.  
:::

::: callout-tip
## Discovering services
Use portals like [FRDR](https://www.frdr-dfdr.ca/repo/) or [Lunaris](https://www.lunaris.ca/) to discover and reuse research data.
:::   

# Data as the new oil 

## Data is a valuable resource{.center}

{{< bi columns >}} Simulation-based approaches, machine learning and artificial intelligence.

{{< bi gear >}} Build workflows based on existing data and pipelines.

{{< bi share-fill >}} Share and reuse data with the scientific community to build collaborations. 


## How do we get there?

- [FAIR]{style="color:green;"} data: Making our data useful to others.
- Adequate and comprehensive [documentation]{style="color:green;"} that contextualizes and describes our data.
- Sharing [reproducible workflows]{style="color:green;"} that ensure the reproducibility of results (protocols, open file formats, analysis workflows).
- Incorporate an [open science culture]{style="color:green;"} into research environments, focusing on transparent reporting of research results and open data sharing.

## Find more supporting material{.center}

-   FRDR [guide](https://www.frdr-dfdr.ca/docs/en/depositing_data/) for deposit research data.
-   Guidance on depositing existing data in [public repositories](https://ethics.gc.ca/eng/depositing_depots.html) 
-   [RDMkit](https://rdmkit.elixir-europe.org/sharing)

::::: {layout-ncol="2" layout-valign="center"}
::::{#first-column}
::: {style="text-align: center;font-size: 50%"}
![](images/frdr-logo-english-first.png){width="90%"}
:::
::::

::::{#second-column}
::: {style="text-align: left;font-size: 80%"}

Visit us at 

https://www.frdr-dfdr.ca/repo/ 

or [contact us](https://www.frdr-dfdr.ca/repo/contactus)

:::
::::
:::::