šŸ§  Coral Market: A self-sovereign back-end for research data-management powered by Ocean Protocol

OceanDAO Round 9
#1

Coral Market: A self-sovereign back-end for research data-management powered by Ocean Protocol

Part 1 - Proposal Submission

Name of Project

Coral Market

Proposal in one sentence

Grant funds are requested to support the development of an open-source application for GDPR-compliant self-sovereign scientific data management and peer-to-peer sharing.

Description of the project and what problem is it solving

Thousands of petabytes of data on human health, economic activity, social dynamics, and scientific observations of the universe and our impact on it is siloed in legacy institutional web infrastructure.

A key challenge for unlocking scientific data is managing controlled access to sensitive data, such as:

  • Personal data provided by a participant in a research study
  • Sensitive data that can be used adversarially or for harm
  • Proprietary data collected at great expense by an investigator

The emergence of peer-to-peer data storage and standards for decentralised identifiers (DIDs) makes it possible to permanently establish a public records archive in a common web infrastructure accessible to all, regardless of professional/academic status, nationality, language, or age, while respecting the intrinsic self-sovereignty of data providers. Furthermore, 250 TBā€™s of data will be free to store on Filecoin, which will accelerate the attractiveness of decentralised storage for scientists to use the Web3 ecosystem. Scientific innovation is one of the most important public goods that need critical funding and the adoption of Web3 by scientists is fundamental if we are to develop an ecosystem with public goods at the forefront of how we build a better web.

This grant will support the development of the middleware for decentralised data storage of large, scientific datasets. This will act as the machine for the Opsciverse which will allow for unified permissions management and peer-to-peer sharing of scientific data, directly lowering the barrier to deployment on Ocean markets. The application will be powered by the Interplanetary File System (IPFS) and Filecoin for content-addressable peer-to-peer data storage, the Ocean Market middleware including but not limited to: Aquarius, Ocean JS lib, PY lib, and Ocean Provider.

Grant Deliverables

  • [ ] Research data object self-descriptions and mapping to Ocean Protocol Data Token Metadata
  • [ ] Architectural specification, diagram, and technical documentation for Coral Market Application Layer
  • [ ] Develop best practices for engineers to onboard onto Web3 Big Data projects
  • [ ] Publish science-friendly onboarding documentation onto Web3 marketplaces
  • [ ] User Research + User Requirement Design Feedback from Scientists
  • [ ] Technical research reports on the integration of Filecoin Storage Auctions with Open Science Data aggregators to provide low-cost storage options for decentralized laboratories
  • [ ] Rebranded Ocean Market fork on Ethereum / Polygon testnet

Which category best describes your project?

Build/improve applications or integrations to Ocean

Which Fundamental Metric best describes your project?

Other - Our primary goal is to spread the adoption of Web3 technology amongst the scientific community. To reflect this goal, we propose the number of scientific research objects that are created on our platform as a new metric. These projects will be composed of datasets, algorithms, models, and other scientific digital objects, some of which are data tokens. We seek to replicate the OceanDAO model for funding scientific projects pre-registered on our platform. Our product directly influences Oceanā€™s ROI through the Ocean Marketplace because our data objects will utilise the ocean ecosystem.

What is the final product?

The Coral Market will be a fully functioning Ocean Market fork with an intrinsic token that is used to create data tokens that correspond to digital research objects such as datasets, pre-computed weights for ML, scientific protocols/experiments. Our team will expand the data token metadata specification to include compatibility with a wide set of scientific self-descriptions following the Open Science Framework. The final product will allow researchers all over the world to pre-register their hypotheses, tokenise their intellectual assets,

How does this project drive value to the ā€œfundamental metricā€ (listed above) and the overall Ocean ecosystem?

The Coral Market is a critical back-end component of the ā€œOpsciverseā€ - a one-stop-shop for scientists to access cloud services for reproducible open science with big data sets. We will use this time and money to develop comprehensive research and specifications for the building we will start in October.

As part of our grant deliverables, we will generate critical feedback for the OceanDAO to understand current data needs and problems from scientistsā€™ perspectives. We currently have a research agreement with Textile, Filecoin and MIT/Dartmouth to identify decentralized file storage challenges and solutions for big data neuroscience laboratories. We are building tools to migrate up to 250TB of neuroscience data unto decentralized file storage, and making this open data directly available to over 20,000 neuroscience researchers around the world. We expect exponential growth of research projects over a period of 5 years beginning in Q1 2022 as we tailor cloud service needs to researchers to scale adoption.

ROI: buck

If awarded, this project will have received a total of 22564 $OCEAN. Our current deliverables for this round include publishing the product specifications of Coral Market for community review, establishing data bridges with scientific research institutions (MIT/Dartmouth), conducting user research with potential scientific data providers and consumers.

ROI: bang

We expect our efforts will provide a template for other teams of researchers and scientists looking to build on Ocean. If we capture 10% of the 800+ neurotech labs we have identified, we can expect them to follow our template to unleash their data. Each research project will include multiple data token objects such as data, models, protocols etc. An example of the average cost of gathering a neuroimaging dataset is ~$800 USD per participant. A research laboratory collects upwards of 100 participants per project (lower bound based on typical statistical power required for neuroimaging sample size analyses). We can expect each dataset to be worth 133.333 $OCEAN at current prices. For example, if 10% of the 800+ labs identified follow our template to contribute data to the Ocean Market, we can expect a total value of 10.666.640 $OCEAN (bang) in neuroimaging data staked on the marketplace.

The hypothetical ROI following this model results in a value of >200 (bang/buck) with a 100% chance of success, >100 with a 50% chance, >40 with a 20% chance, >10 with a 10% chance, >1 with a 1% chance. We believe the chance of success of the realized outcome described above grows significantly based on when in time success is assessed, specifically increasing significantly towards the end of the grantā€™s roadmap for deliverables.

ROI: [bang / buck] x p(success)

Therefore, bang for buck for this proposal can only stand to benefit the Ocean DAO.

Funding Requested

USD$17,500

Proposal Wallet Address

0x33359285f30e7b3386de70ca500f4fe27853765b
(opscientia.eth)

Have you previously received an OceanDAO Grant?

Yes

Team Website

https://opsci.io

Twitter Handle

@opscientia

Discord

Opscientia

Email address

contact@opscientia.com

Current Country of Residence

Opscientia LTD. is a Singapore registered company.

Part 2: Team

Core Team:

Shady El Damaty , M.Sc., Ph.D.

Sarah Hamburg , M.Sc., Ph.D.

Alexandra McCarroll , M.Sc. (in proc)

Liliana Muscarella , M.A.

Fellows:

Kinshuk Kashyap, Fellow

Achintya Kumar, Fellow

Caleb Tuttle, Fellow

Part 3: Proposal Details

Project Deliverables - Category

Software overview:

We will be forking Ocean Market, utilising the current middleware. We will also develop documentation to ensure industry best practices are uniform throughout development. The main aim of this project is to research and architect the backend of the ā€œOpsciverseā€ and carry out user research in order to develop accurate user requirements for our build phase.

Community engagement:

  • Qualitative user research will be conducted with 5+ scientists.

  • Social media (Twitter & Linkedin) will be used to disseminate our message and engage with the wider community.

Project Deliverables - Roadmap

Any prior work completed thus far?

  • Preliminary community outreach has been completed thus far.

What is the project roadmap?

This grant will bootstrap the September phase of our development lifecycle. Articles will be published to the news page of our website to update the community on our happenings.

1393Ɨ717 69 KB

Teamā€™s future plans and intentions

We plan to request funding from OceanDAO to complete our product development pipeline from September to December. We will post updates and links to deliverables for community feedback at the end of each month. Our goal is to build the critical infrastructure for our Open Science research platform, starting with the Web3 back-end.

Additional Information

This grant is a first step in building a decentralised science platform running self-governed, owned, and automated science activities on-chain.

1 Like
#2

@XandraMcC Round 9 Max is $17,500, Round 8 Max was $17,600. I changed to $17,500 in our DAO database, please update it in your proposal.

1 Like
#3

@AlexN - updated, thank you.

#4

Grant Deliverables

Team Opsci is happy to report that we have met all of our Round 9 deliverables and that work continues on our roadmap for Coral, our Open Science Marketplace dApp.

  • [x] Research data object self-descriptions and mapping to Ocean Protocol Data Token Metadata

The Research Object Self Descriptions technical document has outlined basic specifications for a research data object on Ocean with emphases on self description fields that promote interoperability for parent Research Project Objects and their children (Datasets, Experiments, & Notebooks). Next steps include additional research and development, test data token deployments, and interviews with computational scientists to properly assess workflow branch points and requirements.

  • [x] Architectural specification, diagram, and technical documentation for Coral Market Application Layer

The summary of work completed on Coral, the Scientific Data Marketplace, can be found on Opscientia Documentation: Coral Market. We have successfully forked a rebranded marketplace and have published a preliminary architecture for hosting and serving large datasets. Next steps involve updating the front-end with a landing page for scientist-friendly documentation, investigating custom back-end architecture for running our own PoA test network, and support for data storage and computation.

  • [x] Develop best practices for engineers to onboard onto Web3 Big Data projects

Onboarding Documentation for Engineers has been created to help us onboard developers for the next stage of research and development. We expect to bring on contractors for devOps work and data engineering.

  • [x] Publish science-friendly onboarding documentation onto Web3 marketplaces

We have published our community documentation at docs.opsci.io. Next steps include integrating user research and commissioning a front-end designer to build easy to understand onboarding materials for the Coral landing page.

  • [x] User Research + User Requirement Design Feedback from Scientists

We have made our User Survey for Decentralised Research Data Storage & Management results open source for the community. Next steps involve releasing the materials and instrumentation on https://github.com/opscientia/desci/community/surveys and obtaining additional feedback on our first market prototype once it is published and functioning on test net.

  • [x] Technical research reports on the integration of Filecoin Storage Auctions with Open Science Data aggregators to provide low-cost storage options for decentralized laboratories

We have made our research report on Open Neuroscience Data Archival and Retrieval on Filecoin open source. Next steps are to develop tooling to streamline painpoints for pre-processing and preparing data for submission to Filecoin. We are also actively exploring methods for submitting auctions as part of the Ocean.js workflow now that Textile has integrated a Filecoin bridge to Ethereum.

  • [x] Rebranded Ocean Market fork on Ethereum / Polygon testnet

We have deployed a rebranded Ocean Market fork that can be found here https://market-opsci.netlify.app/. We require additional architecture and back-end research to deploy a network on a test net with example scientific assets such as: datasets, computation, experiments, and notebooks.

@AlexN

1 Like