New advances in the development of echinopscis, an extensible notebook for open science on specimens

Nicky Nicolson, Eve J Lucas

Royal Botanic Gardens, Kew

Symposium: “Taxonomy as open science: tool support to facilitate data use for hands-on practitioners”

XX International Botanical Congress

26^th July 2024

• Transitioned from software development into research
• How we can use software development practices in research:
• Reuse
• Automation
• Version control
• Dependency management
• Continuous integration
• Also processes about communication, design & inclusion
• Open science, take-up & how we design & build for participation

​

​

Context: personal & institutional

Image: jesse orrico / unsplash

Proposed early 2000s

​

Systematics Association special issue 2008

​

Since then:

​

• We do everything online

​

• Explosion of data availability

​

• Recognised different roles in research

​

e-taxonomy – moving taxonomic activities online

Image: RBG Kew

Trends in data availability

• Online & remote collaboration
• Awareness of different roles in research
• research software engineer role
• data steward
• Skills development (data & software carpentry)
• Recognition of different activities required for successful research
• Open science: more than open data

​

Wider context: evolving research culture

​

​

• Prototyping tools to enable taxonomic research using specimens
• https://echinopscis.github.io
• An open science experiment, participatory design:
• Local working environment
• Open to choose your workflow
• Access to:
• Specimens
• Names
• Taxonomy
• Literature
• Institutional profiles,
• People
• Built through the Open Life Sciences (OLS) mentoring / ambassador program https://openlifesci.org

​

​

​

e-taxonomy: what tools & practices do we need?

Where to put our efforts?

Browser

Where to put our efforts?

Browser

API

Where to put our efforts?

Browser

API

Something else?

• A personal knowledge manager: for creating & linking research notes
• Emphasises linking
• Data stored locally, using open formats
• Markdown and optional structured data frontmatter
• Works offline
• Extensible architecture – plugins for data access and citation processing
• Active user and developer community

​

…sounds a lot like OpenRefine, which we have adopted with some success

Could this contribute to our management of linked, semi-structured data, as Open Refine has for tabular data?

​

Open by default:

• “Find our project on github”

​

Open by design:

• Our aim is …
• Our roadmap is …
• You can contribute by …
• Submit issues here …
• Expect this kind of behaviour …
• Our decision-making process is …

​

= “Find our project on github” with all these pointers

​

​

​

Open science: what kind of “open”?

Image: Dima Pechurin / unsplash

• Wikidata plugin:
• Consult remote information resource in-context

echinopscis demos (1/5)

• Wikidata plugin:
• Consult remote information resource in-context
• Glossary plugin:
• Highlight a term, get a definition and illustration

echinopscis demos (2/5)

Glossary

• Wikidata plugin:
• Consult remote information resource in-context
• Glossary plugin:
• Highlight a term, get a definition and illustration
• Phylo plugin:
• Define a Newick code block in a document
• Use leaves and internal nodes to search notes

echinopscis demos (3/5)

Phylo

• Wikidata plugin:
• Consult remote information resource in-context
• Glossary plugin:
• Highlight a term, get a definition and illustration
• Phylo plugin:
• Define a Newick code block in a document
• Use leaves and internal nodes to search notes
• “Co-pilot” plugin
• Reformat your data with a large language model assistant
• Text paragraph of specimens to Darwin core format

echinopscis demos (4/5)

You are a diligent scientific research assistant. You transcribe information accurately without including new data for which there is no basis.

​

If you receive the instruction "specimen table" you will try to make a markdown table from my input with the following headers: country, region, locality, latitude, longitude, date, collector, record number, herbarium codes(s).

​

If you receive the instruction "dwc specimen table" you will try to use Darwin Core terms for the table headers where appropriate. Darwin Core is a data standard which defines terms used in biodiversity informatics data sharing.

Co-pilot custom prompt – specimen tables

Co-pilot

• Wikidata plugin:
• Consult remote information resource in-context
• Glossary plugin:
• Highlight a term, get a definition and illustration
• Phylo plugin:
• Define a Newick code block in a document
• Use leaves and internal nodes to search notes
• “Co-pilot” plugin
• Reformat your data with a large language model assistant
• Text paragraph of specimens to Darwin core format
• Text species descriptions to trait matrices

echinopscis demos (4/5)

You are a diligent scientific research assistant. You transcribe information accurately without including new data for which there is no basis.

​

When you receive the instruction "trait table" you will reply "here is your trait table" and provide a table of the descriptive traits included in the species description. For example if you receive a species description with the sentence "stems 3-5mm is diameter, green, pubescent" you will output a table with column headers "character" and "value" and the rows

stem_diameter,3-5mm

stem_color,green,

stem_description,pubsecent

Co-pilot custom prompt – trait extraction

Interested? Checklist of some practical actions

• Review these slides & the demos: https://bit.ly/IBC2024-nicolson
• Look at the echinopscis website & try it out: https://echinopscis.github.io
• Sign up to github: https://github.com/signup

Bonus: Link your github profile to your ORCID

• Participate on the discussion board: https://github.com/orgs/echinopscis/discussions
• Think: if we made a pre-populated echinopscis vault for your group, what data should we include? What kind of collaboration features do you need?

​

Nicolson & Lucas: New advances in the development of echinopscis, an extensible notebook for open science on specimens.

Slides: https://bit.ly/IBC2024-nicolson

Illustrations created by Scriberia with The Turing Way community. Used under a CC-BY 4.0 licence. DOI: 10.5281/zenodo.3332807

Try it out: echinopscis.github.io