Peer Review & Revision (cc-peer-review-revision)

When to trigger

• A decision letter with reviewer reports has arrived (major/minor revision)
• You must triage requested experiments vs. textual changes
• Reviewers ask for additional in vivo / human validation or rigor fixes
• A consultative / cross-referee comment requires reconciling reviews

Cell Press review context

Cancer Cell uses editorial assessment plus external peer review, and Cell Press can run consultative (cross-referee) review where reviewers see each other's comments. Practical implications:

• Reviewers may converge on shared concerns — address the cross-cutting issue, not just each bullet.
• The editor's letter signals which points are essential vs. optional; prioritize those.
• New experiments that strengthen orthogonal validation (in vivo / human) carry the most weight.

Triage every comment

Point-by-point response structure

For each comment:

1. Quote the reviewer's point verbatim.
2. State the change: new data, reanalysis, or text edit.
3. Point to where in the revision it appears (new Figure 4C, Methods p. X, lines Y–Z).
4. If you disagree, give a courteous, data-backed rationale — do not ignore it.

Open the response with a brief summary of the major new experiments and how the paper improved. Keep tone professional and concrete; reviewers and editor read this closely.

Revision discipline

• Do not over-revise into a different paper; address what was asked.
• When adding in vivo / human data, carry through the same rigor (controls, n, randomization, blinding, deposition of any new datasets).
• Update the Key Resources Table, availability statement, and statistics for any new experiments.
• Re-run the cc-submission preflight before resubmitting.
• If you cannot do an experiment, explain why and offer the strongest feasible alternative.

Worked micro-example: a point-by-point entry

Reviewer 2, comment 3 (verbatim): "The MARK7 phenotype is shown only in cell lines. Does loss of MARK7 restrain tumor growth in an immunocompetent in vivo setting?"

Response (the Cancer Cell pattern):

We agree this was the central gap. We generated Mark7-conditional autochthonous tumors and, in a new Figure 5D–F, show that MARK7 loss reduced tumor burden and restored intratumoral CD8⁺ T cell infiltration (n=9 vs. 10 mice; randomized to arms; blinded histology scoring; log-rank p reported in the legend). The in vivo RNA-seq is deposited under GEO [accession] (KRT updated). We have accordingly revised the Summary and Discussion to state a "candidate vulnerability for checkpoint combination" rather than a "therapeutic target," matching the evidence now in hand.

Note the moves: it concedes the point, delivers the highest-value experiment (in vivo, immunocompetent), carries full rigor (n, randomization, blinding), deposits the new data, and narrows the claim in the same breath — the reviewer sees the manuscript improve, not merely defend itself.

Framing pushback that survives consultative review

Because reviewers may see each other's reports, a rebuttal that dismisses one referee can alienate the panel. When you decline an experiment:

• Grant the scientific premise before explaining the constraint (feasibility, timeline, model availability, or that the request tests a different hypothesis).
• Offer the strongest feasible substitute (orthogonal assay, existing cohort reanalysis, published data).
• Where a request is genuinely out of scope, say what the paper does claim and confine the finding accordingly rather than arguing the reviewer is wrong.

Cancer Cell revision failure modes

• Adding an in vivo experiment with a lower bar than the original (no randomization, n of 3, no blinded scoring) — reviewers notice the rigor drop immediately.
• Answering the "does it hold in patients?" comment with more cell-line data.
• Depositing nothing new even though the revision generated sequencing or proteomics data.
• Re-broadening the Summary/Highlights after narrowing the Results — front matter and body must agree.
• A rebuttal letter that resolves minor textual points in detail but leaves the essential mechanistic ask half-answered.

Checklist

• Every reviewer point addressed individually; none skipped
• Essential (editor-flagged) points prioritized and fully resolved
• Cross-cutting concern shared across reviewers identified and answered
• New experiments carry full rigor and n; new datasets deposited
• Each response points to the exact location of the change
• Claims narrowed where evidence did not grow
• Polite, evidence-based rationale for any disagreement
• KRT, statistics, and availability statement updated; preflight re-run

Anti-patterns

• Selectively answering easy comments and ignoring hard ones
• Adding new data with weaker rigor than the original (no controls, undefined n)
• Dismissive or defensive rebuttals to disagreements
• Claiming a change was made without pointing to where
• Re-broadening claims the reviewers asked you to narrow
• Forgetting to deposit datasets generated during revision

Output format

【Decision】major / minor revision
【Essential points】editor-flagged: [...]
【Cross-cutting concern】...
【New experiments planned】(rigor + n + deposition): [...]
【Per-comment plan】R1.1 → ...; R1.2 → ...; R2.1 → ...
【Claims to narrow】...
【Pushback (with rationale)】...
【Next step】execute revisions → re-run cc-submission preflight