PNAS LaTeX Template: Choose the Files and Check the Handoff

An exported archive is a container. The useful question is whether it contains everything required to reproduce the manuscript you intend to send. Fill in the worksheet with your actual filenames; the entries below describe file roles, not mandatory PNAS filenames.

Mark a row unresolved when you cannot supply its evidence. A file can exist without being the correct version, and a PDF can open without containing every intended figure. The worksheet is useful because it makes these separate questions visible.

Keep author approval attached to an identifiable output. If a coauthor reviewed yesterday's PDF while the archive contains today's figure replacement, the source package has moved beyond the reviewed artifact. Rebuild, check the changed material, and obtain the appropriate author review before treating the package as final.

Use a separate copy so the test does not disturb the working project. These are editorial checks for a reproducible handoff, not a claim that all submission systems use the same compiler or accept the same files.

1. Freeze the candidate. Save the PDF you intend colleagues to review and note the project version or export time.
2. Export the source. Keep the original export intact. Use a second extracted copy for the test.
3. Identify the entry point. Confirm which file builds the main paper and which builds the SI. Do not infer this solely from a filename such as main.tex.
4. Rebuild in the receiving environment. Use the compiler and build settings appropriate to the project. Resolve errors explicitly instead of deleting unfamiliar packages until the log becomes quiet.
5. Inspect the output. Compare the first page, author details, equations, figures, citations, references, and final page with the approved PDF.
6. Record differences. Explain each change, or return the package to the author responsible for it. Rebuild again after a repair that changes the output.

If you use Overleaf, consult its guidance on submitting to publishers for the platform's submission workflow. A transfer button and a successful source rebuild answer different questions. Use the journal's current instructions to determine what must actually be uploaded.

Fictional case: A team exports a PNAS project after replacing its main result figure. The working project refers to figures/result-final.pdf, while an older image remains in a local folder used by a collaborator. One author approves a PDF containing the old image; another tests an archive containing the new image. Both PDFs compile successfully.

The failure is approval drift: successful compilation has been mistaken for agreement about the manuscript version. The first diagnostic question is not which PDF looks more polished. It is which figure supports the reported result and which version the authors reviewed.

This example is invented to demonstrate the decision, not drawn from a customer manuscript. The worksheet cannot determine which analysis is scientifically correct. Its job is to expose the disagreement before an author submits the wrong version.

The official SI project supplies a distinct supporting-information template. Its instructions include removing the instructional first page before saving the completed SI PDF. Inspect the final output rather than assuming that replacing the example text removed every instructional surface.

Then check the relationship between documents. Follow each main-paper reference to an SI figure, table, method, or result. Confirm that the cited item exists, carries the intended label, and supports the sentence that points to it. Repeat the check after renumbering or moving an item into the SI.

Do not move essential reasoning out of the main paper just to make the template appear shorter. A complete source package can still contain an incomplete argument. For the broader manuscript decision, use the PNAS submission guide; for the significance argument itself, use the PNAS cover-letter guide.

Start with the earliest relevant error in the log. Later messages may be consequences of the first missing dependency or malformed command. Preserve the unmodified export while testing a correction in your separate copy.

These checks narrow a problem; they are not a universal repair recipe. If the untouched official template fails in your environment, capture the error and environment details for the relevant support team before migrating a full manuscript into it.

The final package should have a known template origin, a successful source rebuild in the environment you tested, an inspected main-paper PDF, checked SI references, and author review of the intended output. Keep unresolved items visible in the handoff note.

Manusights judgment is that version agreement is the strongest final check in this workflow: the source, rendered document, and author decision should describe the same manuscript. Its limit is clear. It does not establish scientific rigor, compliance with every journal instruction, or acceptance.

When the files are stable but the manuscript's claims or limitations remain uncertain, a manuscript evidence review can help examine that separate question. Manusights does not compile your LaTeX project or certify the submission archive.

Sources accessed September 12, 2026.

1. PNAS official template collection, Overleaf: official project selection.
2. NAS research-article template, NAS on Overleaf: current project provenance, class, and layout limitation.
3. NAS single-column mathematics template, NAS on Overleaf: distinct mathematics project.
4. PNAS supporting-information template, NAS on Overleaf: separate SI document and instructional-page removal.
5. Submitting to publishers, Overleaf: platform submission guidance.