AI Genomic Controlled-Access Justification: Drafting dbGaP Access Requests

End-of-lesson check

15 questions · take it digitally for instant feedback at tendril.neural-forge.io/learn/quiz/end-research-ai-genomic-data-controlled-access-justification-r6a3-creators

1. What is the primary function of AI in the dbGaP controlled-access request process?

   1. Generating the final approval decision for data access
   2. Drafting request justifications that align with data-use limitations
   3. Determining whether a research purpose is permitted under the data use terms
   4. Replacing human reviewers on the Data Access Committee

2. A researcher wants to study COPD genetic variants using the COPDGene cohort. The data-use limitation states the data is for 'respiratory disease research only.' The AI drafts a justification mentioning 'lung disease and smoking-related cancer analysis.' Why might the DAC reject this request?

   1. The IRB approval was missing
   2. The request was not formatted correctly
   3. The AI used overly technical language
   4. The request drifted beyond the limitation by analogy (cancer is not strictly respiratory disease)

3. Which organization maintains the data-use ontology that helps align request justifications with dataset restrictions?

   1. The World Health Organization
   2. The National Institutes of Health
   3. GA4GH (Global Alliance for Genomics and Health)
   4. The International Society for Computational Biology

4. What happens after AI drafts a dbGaP controlled-access request?

   1. The request is sent to the Data Access Committee for review and decision
   2. The request is published for public comment
   3. The data is automatically released to the researcher
   4. The request is archived but never reviewed

5. Why is IRB approval referenced in a dbGaP access request?

   1. It is required to use AI for drafting requests
   2. It replaces the need for DAC review
   3. It demonstrates that an ethics board has reviewed and approved the research plan involving human data
   4. It proves the researcher has funding for the project

6. What is a Data Access Committee (DAC) responsible for in the controlled-access data framework?

   1. Writing AI algorithms for data requests
   2. Hosting genomic data on their servers
   3. Training machine learning models on the controlled data
   4. Reviewing requests and deciding whether to grant access based on data-use limitations

7. A researcher requests access to a dataset with the limitation 'non-commercial research only.' The AI drafts a proposal stating the research is for 'academic research purposes.' Why might this still be problematic?

   1. The term 'academic' is not the same as 'non-commercial' and could be interpreted differently
   2. Academic research is always commercial
   3. The AI should have written 'educational purposes' instead
   4. The limitation language was ignored

8. What does 'secondary use' mean in the context of genomic data requests?

   1. Using data for a purpose different from what the original participants consented to
   2. Using data for the second time in a study
   3. Using data that was previously published
   4. Using data that has already been analyzed by another researcher

9. What is the primary reason AI cannot replace DAC reviewer judgment?

   1. AI cannot write complete sentences
   2. AI does not understand genomic science
   3. AI lacks the ability to evaluate whether a specific use falls within the nuance of data-use limitations
   4. AI is not authorized to make legal decisions

10. What type of information is included in a PI institutional certification statement?

    1. Certification that the researcher's institution agrees to the data use terms and will supervise the research
    2. The researcher's publication history
    3. The researcher's salary information
    4. The researcher's educational background

11. What is the COPDGene cohort an example of in this lesson?

    1. An AI algorithm for drafting requests
    2. A data access committee
    3. A data-use ontology
    4. A specific genomic dataset that requires controlled access

12. What is the main risk if AI drafts a request without carefully matching the data-use limitation language?

    1. The researcher will be banned from future access
    2. The request may be rejected for exceeding the permitted scope
    3. The request will be automatically approved
    4. The AI will be turned off

13. In the context of genomic data, what does 'controlled access' refer to?

    1. Data that is encrypted at all times
    2. Data that requires formal request and approval before use, due to ethical or legal restrictions
    3. Data that can only be accessed by doctors
    4. Data that is stored in a secure facility

14. A researcher asks an AI to draft a dbGaP request. The AI produces a document with all required sections. What should the researcher do before submitting?

    1. Submit immediately since AI did all the work
    2. Delete the draft and write from scratch
    3. Review carefully to ensure accuracy and alignment with limitations
    4. Submit to the AI company for approval

15. Which of these is a task AI CAN perform in the dbGaP request process?

    1. Deciding whether to approve the request
    2. Bypassing IRB review
    3. Generating documentation packets matched to DAC requirements
    4. Ignoring data-use limitations to speed up the process