Cell Metabolism's AI Policy: Cell Press Rules for Metabolism and Metabolic Disease Research

Cell Press is part of Elsevier, but the two have slightly different policy structures:

Elsevier's company-wide policy covers all ~2,800 Elsevier journals. It's broadly aligned with COPE and ICMJE: no AI authorship, mandatory disclosure, authors responsible for all content. Elsevier doesn't mandate a specific disclosure location.

Cell Press's specific requirements add structure. The STAR Methods placement is mandatory (not just recommended). Cell Press provides example disclosure language in its author guidelines. And the editorial team actively screens for compliance during the review process.

Cell Metabolism's implementation follows Cell Press without modifications. The journal's Instructions for Authors link directly to the Cell Press AI policy page. There are no journal-specific exceptions or additions.

For practical purposes: if you know the Cell Press policy, you know Cell Metabolism's policy. Don't waste time looking for journal-specific AI rules that don't exist.

Cell Metabolism uses STAR Methods, which has a specific structure. Your AI disclosure goes under Method Details, not in the Key Resources Table, the Resource Availability section, or Acknowledgments.

Good disclosure for a metabolomics paper:

"During the preparation of this manuscript, the authors used ChatGPT (GPT-4, OpenAI) to improve the readability of the Discussion section, particularly the interpretation of metabolite flux data in the context of existing literature. All AI-generated text suggestions were reviewed and revised by the corresponding author (H.Z.) and the senior author (A.K.), who confirmed that the revised text accurately represented the experimental findings. The authors take full responsibility for the content of this article."

Good disclosure for a paper with computational analysis:

"The authors used GitHub Copilot (Microsoft) to assist with writing R scripts for untargeted metabolomics data processing and statistical analysis. All code was independently validated by the bioinformatics specialist (Y.W.) against manual calculations using a subset of the dataset. ChatGPT (GPT-4, OpenAI) was used to improve the language clarity of the Introduction. The authors take full responsibility for the published content."

What wouldn't work:

"AI assisted with manuscript preparation."

Too vague. Doesn't specify the tool, the sections affected, or the scope. Cell Metabolism's editors will send this back with a request for specifics.

The graphical abstract question

Cell Metabolism papers frequently include graphical abstracts, visual summaries that appear in the table of contents and on the journal's website. These are important for visibility and social media sharing.

The rules for graphical abstracts:

• Created with BioRender, Illustrator, PowerPoint, or similar design tools: Fine, no disclosure needed
• Created with AI generative tools (DALL-E, Midjourney): Prohibited
• Created with a hybrid approach (started in Midjourney, refined in Illustrator): Still prohibited, the generative AI origin disqualifies it

If you're unsure, use BioRender. It's purpose-built for scientific illustrations and isn't a generative AI tool. Cell Press editors won't question a BioRender graphical abstract.

The grant-to-paper point is worth flagging. Some researchers use AI to convert text from their grant application into manuscript prose. This is permitted but requires disclosure, the AI is generating new text based on your input, which falls under the manuscript preparation policy.

Cell Press follows a standardized enforcement process:

During peer review:

• Editor contacts corresponding author
• Request to add AI disclosure to STAR Methods
• If the omission appears intentional, the editor may question the integrity of other aspects of the submission
• Severe cases can lead to rejection

After publication:

• Correction: For minor cases (language editing only, didn't affect scientific content)
• Expression of concern: If AI use potentially affected data interpretation or conclusions
• Retraction: If AI generated fabricated data, false claims, or unverifiable content
• COPE referral: For systematic non-disclosure or evidence of fraud

For metabolism research specifically: Many Cell Metabolism papers include metabolomics datasets, flux analyses, and multi-omics integrations that are computationally intensive. If AI assisted with the analysis code and this wasn't disclosed, the reproducibility of the results comes into question. This is a more serious concern than undisclosed language editing because it affects the scientific validity of the findings.

Nature Metabolism is Cell Metabolism's closest competitor. Both publish metabolism research at the highest level. The key policy difference: Nature Metabolism places AI disclosure in a standard Methods section, while Cell Metabolism uses STAR Methods. The substance is the same; the format differs. If you're choosing between the two journals, the AI policy isn't a deciding factor, they're functionally equivalent.

The STAR Methods distinction does affect formatting. Cell Metabolism's structured format means your AI disclosure sits alongside your reagent tables, experimental model descriptions, and quantification methods. At Nature Metabolism, the disclosure is in a free-form Methods section where you have more flexibility in placement and length.

For metabolomics and multi-omics papers:

• If AI helped with data processing code, describe what the code does and confirm it was validated. Cell Metabolism's reviewers often request access to analysis pipelines.
• Don't use AI to identify "interesting" metabolites or pathways in your data. The biological interpretation is where your expertise matters most, and AI-suggested interpretations might not reflect the actual biology.
• If you used AI-assisted tools for pathway enrichment analysis, describe these in your standard Methods as analysis tools, not in the AI disclosure section.

For in vivo metabolism studies:

• Keep AI away from descriptions of animal models, experimental procedures, and physiological measurements. These must accurately reflect what you did in the lab.
• If you used AI to draft the animal ethics statement, verify every detail against your IACUC protocol.

For clinical/translational metabolism papers:

• Same considerations as any clinical journal: don't use AI for patient data, clinical interpretations, or treatment recommendations.
• Metabolic disease papers often include patient cohort descriptions, these should be written by the clinical investigators, not generated by AI.

For all submission types:

• Draft the STAR Methods AI disclosure during writing, not after. It's much easier to remember what you used when you're actively using it.
• Check that your graphical abstract was created with approved tools. This is one of the most common compliance issues.
• Make sure all co-authors know about the AI disclosure. Cell Metabolism papers frequently have 10-20+ co-authors across multiple institutions, don't let the disclosure be a surprise.

Before submission checklist:

• [ ] AI disclosure in STAR Methods → Method Details
• [ ] Tool names, versions, and specific use cases included
• [ ] No AI-generated images or graphical abstract
• [ ] Metabolomics/analysis code validated independently
• [ ] Biological interpretations are human-generated
• [ ] Graphical abstract made with BioRender, Illustrator, or similar
• [ ] All co-authors reviewed the AI disclosure

A free manuscript assessment can help you confirm that your Cell Metabolism submission meets the journal's formatting and ethical requirements before you submit.