Journal of Medical Genetics Submission Guide

Write the central conclusion in five parts: represented population, genetic finding, evidence class, phenotype or mechanism, and clinical consequence. Under each part, name the figure, table, repository record, or analysis that supports it. This produces a compact audit rather than another narrative summary.

For rare-disease work, show how ascertainment affects the apparent phenotype. For cancer genetics, keep germline, somatic, tumour, and family evidence distinct. For functional work, explain why the model tests the human-disease inference rather than only showing that a perturbation changes a molecular readout.

In our editorial analysis, the recurring failure pattern is an inference jump: a variant is novel, rare, or computationally concerning, but the manuscript moves directly to disease causality or clinical action. The useful correction is not more adjective-level certainty. It is a visible chain from ascertainment and phenotype through variant definition, segregation, population evidence, functional validation, alternative explanations, and the exact claim boundary.

For an observational genotype–phenotype study, use the STROBE explanation as a design-and-reporting audit. Trials should check CONSORT 2010, and systematic evidence syntheses should check PRISMA 2020. These are reporting references, not evidence that JMG requires every guideline for every article type; the live author page controls the package.

1. Confirm the live article type and current limits on the official author page.
2. Reconcile nomenclature, genome build, transcript, coordinates, alleles, and variant classification across text, figures, tables, supplement, and repository records.
3. Match the study design to an appropriate reporting guideline through the EQUATOR Network.
4. Align consent, ethics, data availability, controlled-access language, funding, conflicts, and contributor roles.
5. Verify the current submission route from JMG's official page rather than relying on an old bookmark.

BMJ-wide policies support the package, but the journal page controls the current article type and upload requirements. Do not infer JMG rules from a neighbouring BMJ title.

Imagine a study identifies a rare missense variant in several affected relatives and reports a cellular phenotype. The submission is not ready simply because the variant is absent from a local control set. The stronger path separates rarity, segregation, phenotype specificity, computational evidence, functional assay relevance, rescue or orthogonal validation, and remaining uncertainty. If the assay does not model the disease tissue or mechanism, say so and keep the conclusion at the supported classification level.

That boundary should survive the abstract, variant table, figure legend, discussion, and cover letter. A cover letter cannot convert incomplete validation into clinical actionability.

Run this audit after the files are otherwise complete. Late revisions often update one claim while leaving an older, broader version in a legend, table, or letter.

In our analysis of the current official guidance and the manuscript decision implied by it, we found three checks that a generic publisher summary does not resolve. In our editorial analysis, the recurring failure pattern is an inference jump: a variant is novel, rare, or computationally concerning, but the manuscript moves directly to disease causality or clinical action. The useful correction is not more adjective-level certainty. It is a visible chain from ascertainment and phenotype through variant definition, segregation, population evidence, functional validation, alternative explanations, and the exact claim boundary.

Inference jump. Test whether the conclusion moves from an observed result to mechanism, clinical use, or policy without the required comparison and validation.

Population jump. Compare the title, abstract, figure, methods, and conclusion for a change in the represented population, setting, model, or time horizon.

Package contradiction. Reconcile the main claim with the protocol, source files, supplement, data statement, ethics boundary, and cover letter before upload.

For an observational genotype–phenotype study, use the STROBE explanation as a design-and-reporting audit. Trials should check CONSORT 2010, and systematic evidence syntheses should check PRISMA 2020. These are reporting references, not evidence that JMG requires every guideline for every article type; the live author page controls the package.

This source-backed synthesis cannot predict acceptance or editorial priority. It identifies a manuscript-level decision that the publisher instructions do not make for an author.