2013 CTSI Annual Pilot Awards to Improve the Conduct of Research

To facilitate the development, conduct or analysis of clinical & translational research

Sharing Success - Making Open Proposals Self-Serve and Open Source

Proposal Status: 

Abstract. Develop an open source edition of the UCSF Open Proposals software suitable for deployment at external institutions. This will make UCSF's version easier to host and manage, enable other CTSAs to benefit from UCSF's work, and help establish UCSF's thought leadership in the field.

Rationale. In July 2012, CTSI launched UCSF Open Proposals, a web application enabling an open and collaborative community-based proposal submission process. The tool uses a crowdsourcing model to help researchers get valuable input on their proposal before submitting it to the review committee. It helps projects find collaborators, contributors, and advisers, and makes it easy for those with relevant expertise to offer feedback. Finally, it provides a stable platform for ideas even after the opportunity is closed, making it possible for the research community to link to, discuss, and build on previous ideas.

We see four fundamental technical limitations with the current Open Proposals system:

  1. The current version of the Open Proposals application was built in a way that’s very tightly integrated with accelerate.ucsf.edu, CTSI’s services portal. While this allowed us to launch faster, it prevents a standalone deployment, and makes upgrades more difficult.
  2. The design of the system requires the services of a professional web developer to create and configure even the most simple open proposals. This prevents CTSI from making open proposal deployment a turnkey process. At the same time, the system is mature enough at this point that it allows to customize each open proposal forum to fit the initiative-specific needs, at a relatively low cost (but it does require professional resources).
  3. The system is designed in a way where UCSF-specific authentication and identity data access mechanisms are baked directly in, instead of via an open pluggable process. This means that even if the software were to be shared with an outside partner, they would not be able to deploy it at their institution without removing and rewriting UCSF-specific code.
  4. The software is not open source. This prevents UCSF from sharing its work with outside partners, and makes it more difficult to discuss it in venues like the CTSA Toolshop. (At least one major CTSA partner has approached UCSF, wanting to know about the Open Proposals process and software.)

We propose developing a fully standalone open source version of the software, with a flexible authentication and identity backend which can integrate with solutions already in use at UCSF and other institutions. This will allow us to improve usability, maximize national impact, and pay off technical debt.

The solution would fully address limitations 1,2 and 4 and will be a stepping stone on the way to a turnkey solution described in #3.

Plan
In order to convert UCSF Open Proposals to a portable open-source product we need to make the following changes to the application:

  1. Allow system to stand alone by removing dependencies on accelerate.ucsf.edu
  2. Repackage the application to a distributable set of Drupal modules
  3. Implement support for any common network identity provider to support single sign on to the customer’s organization network
  4. Obtain approval from the Office of Technology Management to offer the product under a suitable open source license
  5. Enable external code contributions by hosting the code on an open source version management platform like Github or Bitbucket
  6. Ensure discoverability by listing the product and supporting user documentation on relevant repositories, such as Drupal.org Projects


In parallel, we will be working with several research institutions to find at least one external partner interested in pioneering Open Proposals at their organization, and help initiate a pilot Open proposal opportunity at their institution.

Criteria and metrics for success
This project will be a success if:

  1. An easy to deploy, standalone open source application is made available for the public via the established open source distribution portals
  2. UCSF is able to use the open source application to host its own Open Proposals instance
  3. The application is easy to manage. Application administrator can configure and activate a standard open proposal forum with reasonably low amount of professional assistance from web developer.
  4. An interested partner is found who is willing to pilot an open proposals opportunity at their organization
  5. The partner has installed Open Proposals product and successfully configured a test forum
  6. We receive satisfactory feedback on product’s usability and performance from the partner’s implementation team


Approximate cost and very brief justification
Estimated cost of development: $18,000. This includes anticipated development effort and about 10% contingency.  There may be additional cost of managing communications with the partner (TBD).

Collaborators
Cynthia Piontkowski – Web Producer, User Experience Designer, front end developer for UCSF Open Proposals
Brad Bulger – Drupal and Web Development Expert, Solution Architect, back end developer for UCSF Open Proposals
Anirvan Chatterjee – Solution Architect, Open Source Technology Expert

John Daigre - Communications Director
Oksana Gologorskaya – Product/Project manager, User Experience Designer

 

Comments

This proposal is very cogent, considered, and appropriate, with potential for greatly improving the tool's hosting and use, among other CTSA's and within the broader scientific community.

The "Plan" to implement the proposal is very clearly defined, with specific, seemingly feasable actions needed, at a minimal expense. The "Rationale" presents a robust case for pursing them. I especially appreciate that "Criteria and metrics for success" include satisfactory end-user feedback, from the future partner's implementation team. Partner implementation would be the crucial outcome, so attention to this feedback would benefit future iterations and amplifications.

It's clear that the proposal's authors have very thoroughly analyzed Open Proposals' current requirements and limitations, and future potential. They have produced a very appropriate proposal that would, if implemented, more closely align the tool with the developers' mission: to "enable transparent and collaborative proposal development [and] promote multi-disciplinary team building." I hope this award's reviewers will consider the potential impact this proposal has for optimizing this type of collaboration within the broader community.

 

The Open Proposal platform is an innovative tool that represents an important shift in the paradigm of research funding practices. Early success has demonstrated proof-of-concept, and in order to move the concept forward, there is a clear need to expand access by making the software Open Source. The proposal mentions working with external partners, and this aspect is key. If motivated partners could be identified, perhaps in-kind development support could be provided, distributing the burden of resources and making the decision to support this proposal an obvious YES! 

I support this proposal and would like to collaborate to help with partnership building and outreach to other institutions. As part of the 60-member Clinicial and Translational Science Award (CTSA) consortium that includes the nation's leading academic medical institutions, CTSI is well-positioned to help spread the word about this innovative tool. As Communications Director at CTSI, I'm part of a national network of CTSA communicators, some of whom are already aware of this product and have expressed interest in using it if it becomes available. Additionally, the open-source aspect of this project exemplifies the intent of the CTSA consortium as a venue for sharing solutions, innovative ideas and best practices.

John, thank you for offering  your help! We are happy to have you on the team for this project.

Communicating effectively with the partner institution would be crucial for us to both help the partner succeed with their pilot initiative, and for us to learn as much as possible from their experience with it.

Commenting is closed.