Pnasnexus Writing

Article Type, Structure & Length (pnasnexus-writing)

When to trigger

• You have not chosen an article type, or the type doesn't match the content's length/ambition.
• The manuscript structure is unclear or the Methods are being pushed out of the main text.
• The paper is over the page budget.
• You have not chosen a classification (research area + subcategory).
• The Results read like a lab notebook instead of an argument.

Step 1: choose the article type (this sets the budget)

The page limit counts the whole package — abstract, Significance Statement, main text, figure captions, and table legends all count toward it. There is no separately stated absolute word cap for the 12-page maximum; the "≈4,000 words" figure describes a 6-page article, not the 12-page ceiling. A "6,000-word cap" seen in third-party guides is 待核实 — the official limit is page-based.

Step 2: PNAS Nexus main-text order

A Research Report runs, in order:

1. Title — declarative, specific, parseable by a non-specialist.
2. Author list & affiliations.
3. Significance Statement — 50–120 words (pnasnexus-significance), for Research Reports / Stage 2 Registered Reports.
4. Abstract — ≤250-word single paragraph, no headings (pnasnexus-abstract).
5. Introduction — the gap and why it matters broadly; end on what you found.
6. Results — argument-ordered (below).
7. Discussion — interpretation and broader, cross-field significance. (A combined "Results and Discussion" is acceptable where it reads better.)
8. Materials and Methods — in the main text (see below).
9. Acknowledgments, funding, data availability, references.

PNAS Nexus does not mandate a rigid IMRaD order beyond the in-text-Methods requirement — but the order above is the conventional, reviewer-friendly shape. Confirm any type-specific structure in current guidelines.

Materials and Methods stays in-text (a key requirement)

PNAS Nexus requires a Materials and Methods section in the main text — verbatim: "The main text must contain a materials and methods section which provides enough information for readers to follow the experiments in the paper. This information cannot be provided solely as supporting information." Provide enough method in the body to evaluate and reproduce the work; longer protocols and extended data go to Supporting Information (SI). Do not strip the main-text Methods to chase the page count — move detail to SI instead.

Step 3: classification (required at submission)

PNAS Nexus asks authors to file the paper under one of three major research-area categories, each with subcategories:

• Biological, Health, and Medical Sciences
• Physical Sciences and Engineering
• Social and Political Sciences

Pick the category (and subcategory) that matches the handling and review community you want, consistent with the venue/division call in pnasnexus-fit. Exact subcategory lists are 待核实 — choose from the live list in the submission system.

Length discipline

• If you are over budget, move detail to SI (extended methods, supporting figures/tables, supplementary text) rather than deleting the main-text Methods.
• Remember captions and legends count toward the page budget — tighten them.
• Consider whether the work is really a Brief Report (≤3 pp) rather than an over-padded Research Report.

Structure as argument, not chronology

Order the Results by the logic of the claim, not the order experiments were run:

1. Establish the phenomenon (the headline result).
2. Rule out the obvious alternative explanations.
3. Show the mechanism / generality.
4. Demonstrate the broad, cross-field implication.

Each Results paragraph: claim sentence first, then evidence (figure callout + numbers), then inference.

Cross-references

• Main figures/tables: "Fig. 1", "Table 1".
• Supporting Information: "SI Appendix, Fig. S1 / Table S1" (confirm the exact SI labeling the journal uses); datasets as "Dataset S1".

Output format

【Article type】 Research Report / Brief Report / Registered Report (S1/S2) / Perspective / Review
【Budget】 within page limit for that type? (6 pp pref / 12 max for Research Report) — captions+legends counted?
【Order check】 Title / authors / Significance / Abstract / Intro / Results / Discussion / Materials and Methods / refs — gaps?
【Methods location】 in main text? yes (FIX if only in SI) — extended detail in SI?
【Classification】 research area (Bio-Health-Medical / Physical-Engineering / Social-Political) + subcategory chosen?
【Results order】 argument-ordered, claim-first? yes/no
【Items in main vs SI】 main: [...] / SI: [...]
【Next】 pnasnexus-figures

Anti-patterns

• Do not pick the wrong article type — a single-point study padded into a Research Report wastes the budget; a sprawling study squeezed into a Brief Report loses rigor.
• Do not delete the main-text Materials and Methods to fit the page count — move detail to SI.
• Do not put Methods only in Supporting Information — the journal forbids it.
• Do not omit the classification; it is required at submission.
• Do not forget that captions, legends, abstract, and Significance Statement all count toward the page budget.

Source: brycewang-stanford/Awesome-Journal-Skills → PNAS-Nexus-Skills/skills/pnasnexus-writing/SKILL.md