Research Ethics in Data Intensive Biology: Questions to Ask Before Sharing

If you generate biological data from human samples, behavioral observations, or genomic sequences, the decision to share that data carries ethical weight. You must balance openness with responsibilities to participants, communities, and society. This guide is for principal investigators, data managers, and graduate students who plan to deposit or distribute data from life sciences studies. It walks through the essential questions you need to ask before hitting upload.

Before you share, check your funder's requirements. The NIH Data Management and Sharing Policy mandates that researchers prospectively plan for data sharing, including addressing privacy and consent. Compliance is not optional. However, policy alone does not resolve every ethical tension. You must also consult guidance from organizations like the Committee on Publication Ethics to ensure that your sharing plan respects the rights of participants and the integrity of the research.

At a Glance

The Core Ethical Questions for Data Sharing

Consent: What Did Participants Agree To?

Consent forms written five years ago may not anticipate today's data sharing norms. If your study enrolled human participants, you must confirm that the consent included language about sharing data with external researchers or depositing in public repositories. When consent is absent or ambiguous, you cannot simply share. A study on loss to follow up among adults receiving ART in Ethiopia illustrates how retrospective data sharing can be ethically complex Incidence and predictors of loss to follow up among adults receiving ART in Metekel Zone. The authors used medical records without explicit consent for broad sharing. In such cases, you may need to seek a waiver from your institutional review board or limit sharing to de identified data under a restricted use agreement.

Privacy: Can Data Be Re Identified?

Even after removing names and dates, biological data often retain identifying potential. Genomic sequences, facial photographs, or rare phenotypes can be linked back to individuals. You must assess the risk of re identification. Controlled access repositories, where requesters must sign data use agreements, are safer than open download systems. If your study involves sensitive topics such as death or dying, as in a pilot study on adolescents perceiving an existential intervention A qualitative pilot study on how young adolescents perceive an arts based existential intervention for grappling with the subject of death, full de identification is essential and might require omitting direct quotes that could reveal a participant's identity.

Provenance: Who Made the Data and How?

Provenance means knowing where data came from, who collected it, under what protocols, and with what version of reagents or instruments. Without provenance, sharing is unreliable. Assign persistent researcher identifiers using ORCID to ensure that contributors receive proper credit. Document every processing step. When data are derived from a collaboration, as in the EpiCom clinical trial that brought together patients, healthcare professionals, and industry Co Creation of the EpiCom Clinical Trial, provenance records must clarify which partner generated each dataset and what consent applied.

Reuse Restrictions: What Uses Are Off Limits?

Open data does not mean unrestricted use. You have a right to specify reuse restrictions. For example, commercial exploitation might be prohibited, or data might be limited to non profit research. Write a clear data use agreement. The NIH Data Management and Sharing Policy requires that you state limitations in your sharing plan. If your data come from vulnerable populations, such as children in mental healthcare studies, you should restrict reuse to studies that uphold the same ethical standards. A study protocol on trauma exposed children in Norway Building a collaboration model between primary and secondary mental healthcare levels demonstrates how data sharing agreements can specify that data are only used for approved secondary analyses.

Dual Use: Could Your Data Enable Harm?

Dual use refers to research that can be used for both beneficial and harmful purposes. In biology, pathogen sequences, virulence factors, or host susceptibility data could be misused to engineer bioweapons or to stigmatize populations. While most biological data are low risk, you should consider the dual use potential. For example, data on malaria eradication efforts involving university students in Rwanda University students contribution to advancing malaria eradication efforts in Rwanda could, if mischaracterized, be used to blame certain communities for disease spread. Always frame your sharing with a responsible communication strategy that avoids stigmatizing language.

Responsible Communication: How Will You Report Findings?

Sharing data is not the end. You must also communicate results to participants and the broader public in accessible ways. Engage community stakeholders early. The EpiCom trial example shows how co creation builds trust Co Creation of the EpiCom Clinical Trial. For participants, provide plain language summaries. For journalists, prepare context rich press releases that do not overclaim. The Committee on Publication Ethics offers guidelines on responsible reporting. If your data involve artificial intelligence applications, such as AI facilitating superwood utilization The role of artificial intelligence in facilitating superwood utilization among furniture craft producers, be transparent about how the model was trained and what biases may exist.

Decision Criteria for Responsible Sharing

Before you share, evaluate each criterion against your dataset.

1. Consent adequacy. Does the original consent cover sharing? If no, do you have IRB approval to share or a waiver?
2. De identification level. Have you removed direct and indirect identifiers? Could a motivated adversary re identify individuals?
3. Data sensitivity. Does the data include minors, prisoners, or marginalized groups? Apply stricter controls for sensitive populations.
4. Data use agreement. Have you defined permitted uses, prohibited uses, and attribution requirements?
5. Dual use risk. Does the dataset contain information that could cause harm if misused? Consider a risk assessment.
6. Provenance documentation. Is there a full record of collection, processing, and versioning?
7. Communication plan. Have you prepared a summary for participants and a fact sheet for the public?

If you answer no to any of the above, pause and consult your institution's ethics board before proceeding.

A Practical Workflow for Ethical Data Sharing

Follow this sequence to ensure that ethics are integrated into your data sharing process.

Step 1: Review consent and IRB approval. Retrieve the original consent form and IRB protocol. Confirm that sharing is permitted. If not, submit an amendment or plan for restricted access.

Step 2: Anonymize or de identify the data. Strip direct identifiers. For biological data, consider using a trusted third party to assess re identification risk. Use a formal de identification method approved by your IRB.

Step 3: Write a data management plan (DMP). Include details on data types, formats, metadata standards, and repositories. Align with funder policies, especially the NIH Data Management and Sharing Policy.

Step 4: Define reuse restrictions. Draft a data use agreement that specifies allowed research areas, attribution requirements, and any prohibitions. Use standard templates from repositories like dbGaP or Vivli.

Step 5: Assign persistent identifiers. Register the dataset with a DOI. Assign ORCID to all contributors to link them to the data.

Step 6: Deposit in a controlled access repository. For human data, avoid open repositories. Choose a repository that requires user registration and data use agreements.

Step 7: Prepare communication materials. Write a plain language summary for participants. Prepare a data descriptor or a data note for publication. Work with institutional communications to avoid hype.

Step 8: Monitor reuse. Periodically check who has accessed your data and for what purpose. Report any violations to your IRB.

Common Ethical Mistakes in Data Sharing

Mistake 1: Assuming consent obtained years ago covers modern sharing practices. Consent language from 2010 rarely includes broad open data sharing. Always verify.

Mistake 2: Over de identification. Removing too much data can destroy research utility. Balance privacy with the need for context. For example, removing age and sex may limit replication.

Mistake 3: Ignoring dual use potential. Researchers often dismiss dual use concerns for their own data. Be objective: if your data could be used to target vulnerable groups, restrict access.

Mistake 4: Failing to provide provenance. Sharing raw data without metadata renders it useless. Include instrument settings, software versions, and laboratory protocols.

Mistake 5: Communicating results without participant input. A study on loss to follow up Incidence and predictors of loss to follow up highlights how participants may feel exploited if they learn about findings through the news. Share findings with them first.

Limits and Uncertainties: When Ethics Outruns Policy

Policies often lag behind technology and social expectations. You may find that your consent form does not address whole genome sequencing or that your IRB has no procedures for evaluating dual use of computational models. This is a known gap. The NIH Data Management and Sharing Policy is a framework, but it does not resolve every case. Similarly, the Committee on Publication Ethics offers principles but not step by step guidance for every data type.

Another uncertainty is the long term fate of shared data. Once deposited, you cannot control future uses decades later. Computational methods for re identification improve. A study that seems de identified today might become identifiable tomorrow. You must accept this limitation and build in protections such as term limited access agreements.

Finally, community expectations vary widely. What is acceptable to a participant in one cultural context may be unacceptable in another. A project on malaria eradication involving Rwandan university students University students contribution to advancing malaria eradication efforts illustrates the need for culturally appropriate engagement. There is no one size fits all answer.

Frequently Asked Questions

Q1: Can I share data from deceased participants without consent? Consent for deceased individuals depends on your jurisdiction. Many ethics committees permit sharing if the participant agreed before death or if the data are fully anonymous. Always check local regulations.

Q2: Do I need an ORCID to share data? An ORCID is not strictly required, but it is strongly recommended. Reputable repositories and funders increasingly use persistent identifiers to track contributions. Obtaining an ORCID is free and helps establish provenance.

Q3: What if my data are from a study that involved minors? You need parental consent and often child assent. Sharing of pediatric data requires additional safeguards, such as strict de identification and restricted access. Many repositories have specific policies for children's data.

Q4: How do I handle a data use agreement violation? First, notify your IRB and the repository. Document the violation. You may need to revoke access and, in serious cases, notify the researcher's institution. The Committee on Publication Ethics has flowcharts for handling misconduct.

References and Further Reading

• NIH Data Management and Sharing Policy Official policy and planning guidance for research data.
• ORCID Persistent researcher identifiers and research contributor records.
• Committee on Publication Ethics Publication ethics resources and guidance.
• Incidence and predictors of loss to follow up among adults receiving ART in Metekel Zone, Northwest Ethiopia J Health Popul Nutr
• The role of artificial intelligence in facilitating superwood utilization among furniture craft producers Sci Rep
• A qualitative pilot study on how young adolescents perceive an arts based existential intervention for grappling with the subject of death Discov Ment Health
• Co Creation of the EpiCom Clinical Trial: Bringing the Tuberous Sclerosis Complex Patient Community, Healthcare Professionals, and the Pharmaceutical Industry Together Patient
• University students' contribution to advancing malaria eradication efforts in Rwanda Malar J
• Building a collaboration model between primary and secondary mental healthcare levels to improve care for trauma exposed children in Norway BMJ Open

Related Articles

• Responding to Peer Review: A Structured Revision and Response Process
• Research Collaboration Agreements: Aligning Roles, Data, and Authorship Early
• Scientific Communication for Mixed Audiences
• Research Project Management: A Lightweight System for Complex Studies
• Research Authorship: Discussing Credit, Contributions, and Order Before Submission