際際滷

際際滷Share a Scribd company logo

Thoughts on Biological Data Sustainability

Download as PPTX, PDF
Philip Bourne
Philip Bourne

This document discusses approaches to improving biological data sustainability. It proposes moving from the current BDS 1.0 model to a BDS 2.0 model. BDS 1.0 is characterized by increasing data and costs but decreasing funds for innovation. BDS 2.0 would recognize the monetary value of data and embrace public-private partnerships and a data economy. It suggests a "data credits" system where data curation is a service with monetary value. The document provides examples of how this could work for the Protein Data Bank (PDB) and more globally. It argues BDS 2.0 could encourage competition, globalization, and private sector engagement to better foster sustainable and FAIR biological data.

1 of 11
Thoughts on Biological Data Sustainability For the Global Biodata Consortium Advisory Board Meeting Taken from Terrence R. Johnson & Philip E. Bourne The Biological Data Sustainability Paradox https://arxiv.org/abs/2311.05668 Philip E. Bourne peb6a@virginia.edu /pebourne January 15, 2024
What we propose is not intended as an answer to the biological data sustainability problem, but rather a call to think differently in collaboration with people outside biomedicine who study such problems.
The Current Situation Biological Data Sustainability (BDS) 1.0 More data, more demand, more money spent, less money for innovation Mixture of aggregated and standalone data resources Current culture - sense of entitlement Community expects free access to data Data providers expect to be funded ad infinitum Sense of ownership, sovereignty Hard to see how this scales in a digital economy
Enter Global Biodata Consortium Noble cause Chicken and egg situation with funders: Funders in a wait and see mode; seeing needs more international funders GBC core resources Identify appropriate resources Stimulate discussion and action between them towards sustainability Hard to see how this moves the needle on sustainability, but Not supporting the GBC only makes it worse
Towards Biological Data Sustainability (BDS) 2.0 Recognize that data have monetary value easier in an AI world Recognize the value of public-private partnership (PPP) Embrace a data economy one model based on cap and trade Data has value expressed in some way e.g., credits Data curation as a service has value that can be traded in credits
Consider an Example - PDB BDS 1.0 No PPP Fear of funding loss Common data representation Multiple redundant global sites Curation done in house Competitive feature creep BDS 2.0 PPP Encouraged as part of the model Common data representation Single globally managed site Curation possibly outsourced Features defined by global community
BDS 2.0 for PDB Only - Data Credits and Service Model Global cap on how much will be spent on PDB agreed by funders GBC or funders (broker) create credits in accordance with the cap that can be traded and audited Credits distributed according to current data deposition and curation work loads Sites processing lots of data will use their credits and can request, via the broker, credits from other sites that are less productive Sites can subcontract paying with credits New sites can be allocated startup credits by the broker Sites that are not productive or non-conformant to data standards can be refused credits Private sector heavy data users will be asked to contribute credits via the broker and can specify who gets the credits initially
BDS 2.0 for PDB Only - Data Credits and Service Model Advantages/Disadvantages Advantages Globalizes the enterprise Audits the enterprise Encourages competition (for credits) across sites The broker can impose rules that foster FAIR data across the enterprise Private sector engagement Disadvantages Level of global cooperation not hitherto seen Over-competition might impact collaboration
BDS 2.0 Global vs PDB - Data Credits and Service Model Goes beyond PDB sites Curation is decentralized Introduces data users and data producers (i.e., researchers) into the system for credits Data resource becomes a credit broker Resource awards credits to data depositor when data are downloaded Data user expends credits to download data Researchers without credits can curate data to obtain credits Researchers with too many credits (through deposition) can offer credits to other researchers to curate their data Researchers can buy and sell credits
https://datascience.virginia.edu/people/alex-gates https://datascience.virginia.edu/people/terence-johnson
Thoughts on Biological Data Sustainability

More Related Content

Thoughts on Biological Data Sustainability

  • 1. Thoughts on Biological Data Sustainability For the Global Biodata Consortium Advisory Board Meeting Taken from Terrence R. Johnson & Philip E. Bourne The Biological Data Sustainability Paradox https://arxiv.org/abs/2311.05668 Philip E. Bourne peb6a@virginia.edu /pebourne January 15, 2024
  • 2. What we propose is not intended as an answer to the biological data sustainability problem, but rather a call to think differently in collaboration with people outside biomedicine who study such problems.
  • 3. The Current Situation Biological Data Sustainability (BDS) 1.0 More data, more demand, more money spent, less money for innovation Mixture of aggregated and standalone data resources Current culture - sense of entitlement Community expects free access to data Data providers expect to be funded ad infinitum Sense of ownership, sovereignty Hard to see how this scales in a digital economy
  • 4. Enter Global Biodata Consortium Noble cause Chicken and egg situation with funders: Funders in a wait and see mode; seeing needs more international funders GBC core resources Identify appropriate resources Stimulate discussion and action between them towards sustainability Hard to see how this moves the needle on sustainability, but Not supporting the GBC only makes it worse
  • 5. Towards Biological Data Sustainability (BDS) 2.0 Recognize that data have monetary value easier in an AI world Recognize the value of public-private partnership (PPP) Embrace a data economy one model based on cap and trade Data has value expressed in some way e.g., credits Data curation as a service has value that can be traded in credits
  • 6. Consider an Example - PDB BDS 1.0 No PPP Fear of funding loss Common data representation Multiple redundant global sites Curation done in house Competitive feature creep BDS 2.0 PPP Encouraged as part of the model Common data representation Single globally managed site Curation possibly outsourced Features defined by global community
  • 7. BDS 2.0 for PDB Only - Data Credits and Service Model Global cap on how much will be spent on PDB agreed by funders GBC or funders (broker) create credits in accordance with the cap that can be traded and audited Credits distributed according to current data deposition and curation work loads Sites processing lots of data will use their credits and can request, via the broker, credits from other sites that are less productive Sites can subcontract paying with credits New sites can be allocated startup credits by the broker Sites that are not productive or non-conformant to data standards can be refused credits Private sector heavy data users will be asked to contribute credits via the broker and can specify who gets the credits initially
  • 8. BDS 2.0 for PDB Only - Data Credits and Service Model Advantages/Disadvantages Advantages Globalizes the enterprise Audits the enterprise Encourages competition (for credits) across sites The broker can impose rules that foster FAIR data across the enterprise Private sector engagement Disadvantages Level of global cooperation not hitherto seen Over-competition might impact collaboration
  • 9. BDS 2.0 Global vs PDB - Data Credits and Service Model Goes beyond PDB sites Curation is decentralized Introduces data users and data producers (i.e., researchers) into the system for credits Data resource becomes a credit broker Resource awards credits to data depositor when data are downloaded Data user expends credits to download data Researchers without credits can curate data to obtain credits Researchers with too many credits (through deposition) can offer credits to other researchers to curate their data Researchers can buy and sell credits