U19 Data Science Consortium Survey 2022

Background of survey and first meeting.

  1. Convened representatives from each U19 Data Science Core on July 5th, 2022, to talk about what could best be addressed by a collective effort. Topics covered were:
  • Do you have a success story to share?
  • What collaborations have you formed with other U19 BCPs?
  • What is the prototype framework (sandbox/pilot) you are creating (e.g., workflows, analytical/computing resources)?
  • What domains are you reaching?
  • What are the outcomes?
  • What are the challenges/barriers?
  • What are your next steps?
  1. Collected responses of Data Science Core activities from each U19 in the following document:
  1. Recognized development efforts can be leveraged, shared with lessons learned.
  2. We decided to conduct a survey across U19s to learn which tools, technologies, and methods are used across our collaborations, and identify any shared difficulties and challenges that could best be addressed by a collective effort. Manuel Schottdorf is making the survey. Edgar Walker + Guoqiang Yu support.
  3. Data was collected between November 2022 and January 2023.

Overview of the results

These are links to the survey, and summary slides, together with further information produced during the survey meeting on January 27th, 2023. This was presented to all U19 data science core leads and PIs.

  1. Important discussion points that came up:
    1. Need to make it more rewarding for trainees to engage in the data science activities. Their contributions are not as measurable as with a traditional paper. "There is no reward in constructing comprehensive pipelines or infrastructure."
    2. What are the career paths for data scientists in neuroscience? Big turnaround, even between July 2022 and January 2023 meetings. Some U19s' data science core personnel have left science.
    3. How can we recruit data scientists to come to neuroscience?
    4. How can we incentivize better data science in neuroscience?
    5. Use the "depressing" state-of-art, documented in the survey, to motivate future improvements.
    6. It is hard to change existing lab "culture" - instruments, pipeline, analysis.
    7. Ramp-up of data scientists when first introduced into the lab
    8. Moving from incremental activities to true culture change
    9. Advertise for citations to digital instruments/objects/pipelines as a first step towards measuring contributions.
    10. Buy-in by U19 leadership to have an "Analysis First" practice.
  2. Follow up activities:
    1. Who attended July 2022 Consortium Meeting?
    2. Who attended this January 27, 2023 Consortium Meeting?
    3. Understand why non-responders didn't contribute (Osmonauts, OXT, FlyLoops, MSCZ).
    4. Data science trainee/career issues.
    5. Neuroscience living in the present - not planning for the future; excited about doing the experiment now, but not about using data in a distant future.
    6. Not motivated to go back to "old" data.
    7. Need to understand better the interactions between the data scientists and the leadership as well as the trainees doing the experiments.
  3. Plan:
    1. Present this to a broader audience with trainees in June.
    2. Include a Workshop to gather ideas.
    3. Write all of that down.

Presentations

Talk based on Manuel's, Edgar's and Guoqiang's survey for Creating Best Practices for Experiment & Theory Integration, Data Sharing, and Training at the 2023 BRAIN Investigator meeting. Presentation summarizing and synthesizing survey results on June 11, 2023.

Workshop on the next day, June 12, 2023. Feedback from the community in work groups.

Neuroscience Gateway Satellite Workshop at the SfN 2023 meeting, given at the NIH on Nov 11, 2023.

  • Slide deck (11/2023)
  • Key message: We need to write up our results, and publish the findings.

 

Publications

Preprint from Manuel, Guoqiang and Edgar (March 2024) with all references

  • Schottdorf M, Yu G, Walker EY. Data science and its future in large neuroscience collaborations. bioRxiv 2024 Mar 25:2024.03.20.585936. doi: 10.1101/2024.03.20.585936. PMID: 38585895; PMCID: PMC10996530. Link

Streamlined & peer-reviewed version:

  • Schottdorf M, Yu G, Walker EY: Data science and its future in large neuroscience collaborations. Neuron 112(18): 3007—3012 (2024). Link

 

Return to the Main U19 page

Table sorting checkbox
Off