From Preprint to Published: The Secret Life of Scientific Manuscripts

A stack of organized documents and folders secured with colored clips, showcasing a mix of papers and binders in a workspace.

What really happens to research preprints after they hit the digital world? At GESIS – Leibniz Institute for the Social Sciences, a dedicated team has documented the journey of over 145,000 bioRxiv preprints, revealing surprising twists and the evolving relationship between speedy online sharing and formal journal publication. The newly released PreprintToPaper dataset shines a light on these hidden scientific stories—but just how many make it all the way to print, and what changes along the way?

Was passiert wirklich mit Forschungsvorabdrucken, nachdem sie in der digitalen Welt veröffentlicht wurden? Am GESIS – Leibniz-Institut für Sozialwissenschaften hat ein engagiertes Team den Weg von über 145.000 bioRxiv-Preprints dokumentiert und dabei die sich wandelnde Beziehung zwischen schnellem Online-Sharing und formeller Veröffentlichung in Fachzeitschriften festgehalten. Der neu veröffentlichte Datensatz PreprintToPaper bildet die Grundlage für Antworten auf damit verbundene Fragen: Wie viele Preprints schaffen es tatsächlich bis zur Veröffentlichung, und was passiert auf dem Weg dorthin?

DOI: 10.34879/gesisblog.2026.116

A Snapshot at the Digital Frontier

Picture it: a buzzing research lab at the height of the COVID-19 pandemic. Scientists, racing against time, upload their results to preprint servers, ensuring the world can read about their discoveries weeks—or even months—before journals approve the final version. Preprints act as the lifeblood of rapid science, especially when speed can mean saving lives. But what becomes of these early manuscripts? Do they evolve, or simply fade away in the rush of new information?

Opening the Black Box: Why Track Preprints?

Why should we care about these digital drafts? The rise of preprints—unreviewed, early versions of scientific manuscripts—has transformed scientific communication. In the life sciences, preprints have accelerated the spread of ideas and spurred collaborations, especially during the pandemic. Yet, whether (and how) these preprints morph into peer-reviewed, published articles has remained a mystery.

The PreprintToPaper dataset, created by Fidan Badalova, Julian Sienkiewicz, and Philipp Mayr at GESIS, offers a large-scale dataset/map of this journey. Drawing data from over 145,000 preprints submitted between 2016–2018 and 2020–2022, it links each bioRxiv preprint to its journal “destination”—or notes where the trail goes cold.

Tracking the Preprint Lifecycle

So, what does the scientific “migration” look like? The research team assembled meticulous metadata for every preprint: titles, abstracts, author lists, institutions, subject categories, submission and publication dates—even the number of manuscript versions. Each preprint fell into one of three categories: published (formally linked to a journal article), preprint-only (never published in a journal), or “gray zone” (possibly published, but lacking a clear journal link).

Imagine a digital chessboard: some preprints smoothly glide toward shiny journal covers; others linger on the periphery, altered through multiple versions, awaiting their fate. To enhance accuracy, the researchers computed similarity scores for titles and authors, and even created a human-annotated subset of thorny “gray zone” cases, confirming which preprints had really made the leap into journal publication.

What the Numbers Reveal

Are preprints just stepping stones, or digital dead-ends? The numbers tell a striking story: in the dataset, 90,614 out of 145,517 preprints (across both periods) could be confidently linked to journal articles. But that’s not the whole story—by using sophisticated matching methods, the team uncovered thousands more “gray zone” cases, where minor tweaks to the title or author lineup meant a formal publication link was missing.

Perhaps most intriguingly, the fraction of preprints making it all the way from digital draft to journal article seemed to drop sharply between 2016 and 2022—unless gray zone matches were included. Then, the perceived decline nearly vanished. In the age of information overload and “paper tsunamis,” these subtle shifts matter.

What Makes a Preprint ‘Stick’?

Do factors like author count, title length, or even writing style influence a preprint’s next steps? With the rich, nuanced metadata from PreprintToPaper, future researchers can probe not just whether a preprint is published, but why. Is it the scientific content that matters, or the clarity and polish of the presentation? And as natural language processing tools become more powerful, scholars can track how preprints change across versions—detecting everything from major rewrites to subtle tweaks in author order.

A Dataset for the Future

The PreprintToPaper dataset, available openly through Zenodo, isn’t just a static list—it’s a launchpad for deeper explorations. Whether you’re studying information overload, open science, or the nuts and bolts of scientific communication, this resource opens new doors.

Will preprints continue to surge, or be tamed by evolving peer-review norms and better linking systems? How many digital manuscripts will remain forever in limbo? The story is still unfolding—and with PreprintToPaper, the next chapter is ready to be written.

Original scientific publication:
Badalova, Fidan, Julian Sienkiewicz, and Philipp Mayr. 2026 (Forthcoming). “PreprintToPaper dataset: connecting bioRxiv preprints with journal publications.” Scientific Data 13.
DOI: https://doi.org/10.1038/s41597-026-06867-3

Dataset:
DOI: https://doi.org/10.5281/zenodo.17183276

This article was written by Christian Kolle with the support of ChatGPT 4.1 based on the original scientific publication and reviewed by one of the researchers involved, Fidan Badalova.