# Journal Data Sharing Policies: What Top Life Science Journals Require Top life science journals differ in one decisive way: some require you to deposit your data in a public repository before they will accept your paper, while others only ask for a written statement saying where the data are and how to get them. The rule of thumb is simple. If your data type has a dedicated community archive, such as a sequence read archive for raw sequencing reads, plan to deposit there and choose a journal whose policy matches that expectation. If your data are sensitive, human-derived, or hard to move, look for a journal that accepts a well-documented availability statement or controlled access route. This guide walks through the main journal data sharing policies in the life sciences, the options you will encounter, the pros and cons of each, how to compare data repositories, and the mistakes that most often slow down submission or quietly undermine a published paper. ## What Journal Data Sharing Policies Actually Ask of You ![Decision path asking whether a domain archive exists, whether data are human and consent limited, and which journal policy route to take.](/article-images/illustrations/journal-data-sharing-policies-life-sciences-policy-triage.webp) A data sharing policy is a journal's formal statement about what research data must be accessible when a paper is published. Most policies in the life sciences combine four elements, and understanding each one prevents most submission problems. **Scope.** The policy names what counts as data. Raw instrument output, processed results, code, and protocols are treated differently. A sequencing paper usually must share raw reads, while a [clinical trial protocol](/blog/guides/clinical-trial-protocol-development-key-components-and-best-practices) paper may only need to state that de-identified participant data will be shared on request. **Timing.** Some journals require deposition at submission, others at revision, and others only at acceptance. Journals such as PLOS One and BMJ Open ask for a data availability statement during submission, which forces the decision early rather than at proof stage. **Enforceability.** A policy can be enforced by editors who check accession numbers, or it can rely entirely on author good faith. This distinction matters more than most researchers expect, because verification effort varies enormously between journals. **Exceptions.** Every policy has carve-outs for ethical, legal, or proprietary restrictions. Human genomic data, licensed clinical datasets, and third-party data usually qualify. The policy still expects you to describe the restriction and, where possible, point to a controlled access route. Underlying all of these is the FAIR framing, which stands for findable, accessible, interoperable, and reusable. Recent work in microbiome research proposes tiered standards for FAIR sequence data and metadata sharing precisely because written policies alone have not produced consistent practice [1]. Standardized frameworks such as STARDIT, the Standardised Data on Initiatives project, extend the same logic to reporting who contributed to a project, what tasks they performed, and what impacts were observed, in a machine-readable format across disciplines [2]. ## The Five Main Policy Options Most journal data sharing policies in the life sciences fall into one of five categories. Each option is explained here on its own terms, with its practical consequences. ### Option 1: Mandatory Deposit in a Public Repository This is the strongest policy. The journal requires that specified data types, usually raw data underlying the figures, be deposited in a recognized public repository before publication. Editors may verify accession numbers or digital object identifiers during review. For sequence-based research, this model dominates. Domain archives operated by organizations such as NCBI and EMBL-EBI accept raw reads, assemblies, and expression data at no cost to the depositor. Journals publishing microbiome, genomics, and structural biology work routinely require accession numbers for these archives. The pros are clear. Deposited data carry a permanent identifier, remain accessible even if a lab moves or a link breaks, and allow independent reanalysis. The cons are effort and exposure. You must prepare metadata to community standards, and competitors can see and reuse your dataset the moment it becomes public. Some repositories allow an embargo until publication, which softens the timing problem. ### Option 2: Data Availability Statement With Deposit Encouraged This is the most common policy in broad-scope life science journals. The journal requires a data availability statement in the manuscript but does not force deposit in a specific archive. Typical statement options include data deposited in a named repository, data included in the paper or supplements, data available from the corresponding author, third-party data with a citation, and restrictions explained on ethical or legal grounds. The advantage is flexibility. You can handle sensitive datasets, large collaborations, and proprietary constraints without blocking publication. The disadvantage is weak enforceability. A systematic evaluation of 2,929 human gut microbiome publications, plus case studies in soil and gut microbiota, found that nearly half did not meet minimum standards for sequence data availability, and the study's authors concluded that sequence data availability statements are often ineffective in practice [1]. A statement saying data are available on request does not guarantee that a reader five years later receives anything. ### Option 3: Data as Supplementary Materials Some journals accept datasets uploaded alongside the manuscript as supplementary files. This is common for small tables, flow cytometry panels, machine-readable versions of figures, and spreadsheets that reviewers need during evaluation. The pros are simplicity and immediacy. Reviewers see the exact files, and no third-party account is needed. The cons are significant for anything large or complex. Supplementary files often lack [persistent identifiers](/knowledge/bioinformatics/persistent-identifiers-in-bioinformatics-selection-criteria-and-implementation-workflow), are not indexed for search, and become effectively unusable when the host journal website changes. Metadata standards are rarely enforced at this route, which creates the harmonization problems documented across microbiome studies, where poor metadata standardization creates a high barrier to cross-study comparison [1]. A sensible compromise used by many authors is to place the file readers actually need in the supplements and deposit the full raw dataset in a repository, linking the two. ### Option 4: Controlled or Managed Access For sensitive human data, the policy option is not open deposit but controlled access through a certified archive. The data are stored centrally, and qualified researchers apply for access through a data access committee governed by consent terms and, where applicable, privacy law. This model is standard in precision oncology and clinical genomics. The Lancet Oncology Commission on Cancer Genomics and Precision Oncology identified robust data stewardship, including effective data sharing and global integration, as a central requirement for precision oncology to deliver reproducible and equitable benefit, warning that without it the field stays fragmented across projects, platforms, and jurisdictions [3]. [Newborn screening](/knowledge/molecular-biology/newborn-screening) genomics faces the same governance question, with reviews of the field explicitly addressing global policy variation and data-sharing governance frameworks as multi-omic screening expands [4]. The pros are ethics compliance and feasibility for consent-limited data. The cons are slower reuse, because access applications take time, and uneven policies between countries and repositories, which limits the global integration that large-scale genomics needs [3]. ### Option 5: Field Norms and Standardized Reporting Frameworks Some fields have no single journal mandate but strong community standards enforced through peer expectations and structured frameworks. In microbiome research, a proposed tiered badge system grades compliance for sequence data and metadata, and an automated evaluation tool can assess adherence in published amplicon and metagenome papers [1]. In reporting terms, newer checklists extend the same idea to emerging data types, for example the TRIPOD-LLM checklist used to structure reporting for a large language model decision aid evaluation in pediatric rheumatology [5]. Beyond biomedicine, STARDIT provides a standardized way to report initiative-level data, including who was involved and what impacts occurred [2], and public health work on emerging food business models has separately called for improved data sharing and cross-sector collaboration as a policy need [6]. The pro of this option is that standards evolve with the science. The con is inconsistency, because two journals in the same field can apply the same framework at very different levels of rigor. ## Side-by-Side Comparison of the Policy Options | Policy option | What you must actually do | Typical timing | Enforceability | Best suited for | Main drawback | |--|--|--|--|--|--| | Mandatory public deposit | Deposit raw data in a named community archive and provide accession numbers or DOIs | Submission or revision | High, editors can verify identifiers | Sequence data, structures, expression datasets | Metadata preparation effort, immediate public visibility | | Availability statement | Write a specific statement naming the repository, the accession, or the access route | Submission | Low, relies on author follow-through | Mixed data types, collaborative projects | Statements are frequently not fulfilled in practice [1] | | Supplementary materials | Upload curated files with the manuscript | Submission | Medium, reviewers see the files | Small datasets, tables readers need for evaluation | No persistent identifier, poor reusability | | Controlled access | Deposit in a managed access archive and describe the application route | Submission or acceptance | Medium to high, governed by the archive | Human genomics, clinical data with consent limits | Slow reuse, uneven governance across jurisdictions [4] | | Framework and norms based | Follow field-specific tiers, badges, or checklists | Varies by venue | Varies, often peer-driven | Fast-moving fields, new data modalities [5] | Inconsistent application between journals | ## Comparing the Options on the Criteria That Decide Real Choices ### Enforceability and Verification ![Comparison of mandatory deposit, availability statement and controlled access on verification, reuse, timing and best use.](/article-images/illustrations/journal-data-sharing-policies-life-sciences-statement-vs-deposit.webp) This is where policies separate most sharply. A mandatory deposit policy can be checked in minutes, since an editor either sees a working accession number or does not. An availability statement cannot be checked at all at submission, which is exactly why the microbiome evaluation found that availability statements, as commonly written, fail to deliver accessible data in a large fraction of published work [1]. When choosing among journals, the practical question is not whether the journal has a policy but whether anyone verifies it. ### Reuse Value Deposit beats statement, and deposit with standardized metadata beats deposit without it. The microbiome assessment found that even where sequence data are shared, inconsistent metadata reporting blocks harmonization and cross-study comparison [1]. If your goal is a dataset other scientists can actually combine with theirs, budget the metadata effort. Standardized, machine-readable reporting frameworks exist for this reason [2]. ### Ethics and Governance Fit Human-derived data rarely fit open deposit. Controlled access preserves privacy while keeping the data discoverable, and it is the route assumed by governance frameworks in clinical genomics and population-scale screening programs [4][3]. Choosing a journal whose policy recognizes controlled access as a valid route saves you from a late revision crisis. ### Cost and Effort Domain archives for sequences, structures, and expression data are free and streamlined, so the real cost is your time in formatting metadata. Generalist repositories accept almost any file type and issue a persistent identifier, which suits data types without a dedicated archive. Controlled access routes add an application review layer on the user side, which is the deliberate trade for privacy. ### Alignment With Funders and Reviewers Most life science funders expect research data sharing in some form, and journal policies increasingly mirror this. Reviewers in data-intensive fields increasingly ask where the data live before they ask anything else. A journal with a verifiable deposit policy removes that negotiation from review. ## Data Repository Comparison Picking the repository matters as much as picking the journal, because the repository determines whether your data are findable, how they are licensed, and whether reviewers can check them. Most options fall into four types. | Repository type | Examples and operators | Best for | Access model | Key strength | |--|--|--|--|--| | Nucleotide and read archives | GenBank and SRA at NCBI, ENA at EMBL-EBI | Raw sequencing reads, assemblies, submissions tied to journal policy | Open after any embargo | Free, community-mandated, permanent accession numbers | | Expression and [functional genomics](/blog/guides/functional-genomics) archives | GEO at NCBI | [Gene expression](/blog/guides/gene-expression) and epigenomic datasets | Open | Structured metadata designed for reuse | | Generalist repositories | Institutional and disciplinary generalist archives | File types without a domain archive, code, protocols, mixed outputs | Open, often with DOI minting | Accepts nearly anything, flexible licensing | | Controlled-access human data archives | Managed access archives for genomic and clinical data | Consent-limited human genomics | Application through a data access committee | Privacy compliance with discoverability | Two selection rules follow from this table. First, if a domain archive exists for your data type, deposit there, because journal policies and peer reviewers expect those accession numbers and the archives are built for reuse. Second, if no domain archive fits, choose a generalist repository that issues a persistent identifier and supports structured metadata, since a bare link to a lab server satisfies almost no current policy and decays quickly. ## Worked Scenarios ### Scenario 1: An Amplicon Microbiome Paper You have 16S rRNA amplicon data from a gut microbiome study. The strongest route is to deposit raw reads in a sequence read archive with sample metadata following community standards, cite the accession in the availability statement, and aim for the highest tier in the emerging badge framework rather than the minimum [1]. This matters because the evaluation of published microbiome work showed that roughly half of studies miss even minimum data availability standards and that metadata gaps block reuse even where reads are public [1]. Depositing well above the minimum turns your dataset into a citable, reusable asset rather than a compliance chore. ### Scenario 2: A Multi-Institution Cancer Genomics Study You lead a precision oncology consortium with tumor and matched normal sequencing from consenting patients. Open deposit is generally not appropriate, so the plan is controlled access through a recognized managed access archive, a clear description of the access route in the availability statement, and investment in the data architecture that lets collaborating institutions integrate results. The Lancet Oncology Commission framing applies directly here, since fragmented data across projects and jurisdictions is the stated failure mode that good stewardship must prevent [3]. The newborn screening literature raises the same governance issues as multi-omic programs scale globally, including the genomic data gap that can worsen health disparities [4]. ### Scenario 3: Real-Time Public Health Modeling Your group produces weekly pathogen transmission estimates and mortality forecasts during an outbreak. During the COVID-19 response, one modeling team published near real-time weekly transmissibility estimates and death forecasts for every country with sustained transmission, shared online continuously, and later evaluated those forecasts retrospectively across 81 countries [7]. The lesson for data policy is that continuous public sharing of model outputs, with methods documented, is compatible with high-visibility publication and is often expected when results inform policy in real time. ## Pros and Cons Summary **Mandatory deposit.** Strongest for reproducibility and verification. Costs metadata preparation time and removes timing flexibility. **Availability statement.** Maximum flexibility for unusual or sensitive datasets. Weakest enforceability, with documented failures at scale in data-intensive fields. **Supplementary materials.** Fast and reviewer friendly. Weak persistence and no real metadata structure. **Controlled access.** The only viable route for most consent-limited human data. Slower reuse and governance overhead [3]. **Framework based norms.** Adaptable to new data types such as AI model evaluations [5]. Uneven rigor across venues. ## Common Mistakes and Limitations **Treating the availability statement as the finish line.** The single best-documented failure in current practice is the statement that promises data on request and delivers nothing. Evidence from nearly 3,000 microbiome publications shows exactly this pattern [1]. Name a repository and an identifier whenever one exists. **Depositing raw data without metadata.** Data without structured sample metadata cannot be harmonized across studies, which defeats the main purpose of sharing [1]. Budget as much time for metadata as for the deposit itself. **Choosing the repository last.** Authors often pick a journal first and discover at revision that the journal demands an archive they cannot use. Decide the repository route when the dataset exists, not when proofs arrive. **Assuming controlled access equals no sharing.** Many authors avoid any deposit for human data, then write a vague statement that weakens the paper. Controlled access archives exist precisely so sensitive data can be discoverable and governed [4][3]. **Ignoring supplementary file decay.** Large spreadsheets posted only as supplements become unusable when journal platforms change. Pair supplements with a repository deposit. **Overlooking reporting frameworks for new data types.** Work involving machine learning models has its own structured reporting expectations, and following the relevant checklist from the start prevents rewrites. **A note on limitations.** Journal policies change frequently, and this article describes the general structure of current policies rather than the wording of any single journal at any single date. Always read the author guidelines of your specific target journal before submission, and treat this guide as preparation for that step, not a substitute for it. ## Frequently Asked Questions ### Do all top life science journals require public data deposit? No. Requirements range from mandatory repository deposit for specific data types to a simple availability statement with no verification. Data-intensive fields such as genomics and microbiome research lean toward mandatory deposit, while many clinical and social science venues accept statements describing restrictions. ### What is a data availability statement? It is a short manuscript section declaring where the data supporting the paper can be found. A strong statement names the repository and accession number, or explains a specific restriction and the access route, rather than vaguely promising data on request. ### Which repository should I choose for my data? Use the domain archive for your data type if one exists, such as a sequence read archive for raw sequencing reads, because journals and reviewers expect those identifiers. If no domain archive fits, use a generalist repository that issues a persistent identifier and supports structured metadata. ### Can human data ever be shared openly? Usually not without restrictions. Consent terms and privacy rules mean human genomic and clinical data typically go through controlled access repositories, where researchers apply for access through a governed process, and the journal accepts a description of that route as compliance. ### When must data be deposited, at submission or at acceptance? It depends on the journal. Some verify accession numbers at submission, others at revision or acceptance, and many repositories support an embargo that releases the data when the paper publishes. Check the target journal's stated timing and plan the deposit before the final revision. ### What happens if I do not comply with a journal's data policy? At minimum, the manuscript is returned or held at revision until you provide a compliant statement or accession numbers. After publication, noncompliance is harder to fix and contributes to the documented pattern where published availability statements do not result in accessible data [1]. ## Sources 1. [Tier-based standards for FAIR sequence data and metadata sharing in microbiome research.](https://pubmed.ncbi.nlm.nih.gov/40823805/) 2. [Standardised data on initiatives-STARDIT: Beta version.](https://pubmed.ncbi.nlm.nih.gov/35854364/) 3. [Cancer genomic data systems in precision oncology: from integration to clinical implementation.](https://pubmed.ncbi.nlm.nih.gov/42791163/) 4. [The new frontier of newborn screening: integrating genomics, metabolomics, and data science for inborn errors of metabolism.](https://pubmed.ncbi.nlm.nih.gov/42919311/) 5. [An LLM-based decision aid for collaborative decision-making in juvenile idiopathic arthritis: A multimethod evaluation.](https://pubmed.ncbi.nlm.nih.gov/42911979/) 6. [Navigating dark kitchens in England: environmental health, planning, and public health perspectives.](https://pubmed.ncbi.nlm.nih.gov/42786639/) 7. [Retrospective evaluation of real-time estimates of global COVID-19 transmission trends and mortality forecasts.](https://pubmed.ncbi.nlm.nih.gov/37851661/)