Summarize Paper logo SummarizePaper.com

Guides

What did I miss? The close papers your draft does not cite

By Quentin Kral Updated

The referee who writes "the authors should discuss the work of…" has usually found a paper close to yours that your draft does not cite, often because it uses other words for the same thing. What did I miss? (Premium) reads your draft and its references, searches for the closest papers it does not cite, and says for each one whether a referee would expect it, how to position yourself, and the sentence of your draft where to cite it.

To show it on a real case, we gave it a paper as if it were a draft: Bergez-Casalou & Kral (2024), on planetary gaps in the gas of debris discs, published in Astronomy & Astrophysics under a CC BY license. The screenshots show the search of 10 November 2026.

1. Give your draft

Open What did I miss? and drop the PDF of your draft, or paste its text: the title, abstract, introduction and references are enough. Add its .bib if you have one: every reference is then recognised exactly. You can also have the same .bib checked by the bibliography check.

The form: the PDF of the draft selected, an optional .bib, and the box Keep the text of my draft, encrypted, to search again later, ticked.
The PDF of the draft, and the box to keep its text to search again before resubmitting.

Tick Keep the text of my draft to search again later without sending it again: it is stored encrypted and can be deleted at any time. Otherwise it is read for the searches and not stored. Our search took two minutes.

2. What was searched

The result starts with what was searched. For our paper: 114 references read in the draft, 65 of them found in the citation graph; 319 papers found by the searches, 31 of which the draft already cites. The queries, written from the question, method and data of the draft, are under The queries we ran.

What we searched: 319 papers found, 31 already cited by the draft, 114 references read, 65 found in the citation graph; then the export buttons BibTeX, RIS, Open in Overleaf and Send to Zotero, and Save these papers in a collection.
What the search read, and the exports of its papers.

Besides the searches, the candidates include the papers that cite several of your references, which finds work written with other words, and the founding papers the closest candidates cite most often. Retracted papers, the draft itself and the papers it cites are left out. A reference printed without a title is matched by its first author and year: here 19 papers were set apart as probably cited already, to check rather than to add.

3. The papers to cite

The papers are grouped by how much a referee would expect them: Should be cited (8 here), Worth a look (20) and Related (2). Each one says what kind of relation it has with the draft (an earlier result, a data set or tool, a competing method, the same question, a review), how many of your references it cites, then:

Should be cited (8): Dong, Zhu and Whitney (2015), on observational signatures of planets in protoplanetary disks, with how to position yourself, the sentence of the introduction where to cite it, and the Added, Already cited and Not relevant buttons.
The first paper to cite, with the sentence of the introduction where it fits.

Many of the papers proposed here study gaps in protoplanetary discs: the draft compares them with debris discs, and the advice is to cite them as earlier work while saying what differs. Others are close in their data: the vertical structure of gas and dust in four debris discs, to cite where the draft sets up its radiative transfer.

Worthen et al. (2023), Vertical structure of gas and dust in four debris disks, marked Data or tool, citing 15 of the draft's references, to cite in section 2.3 Radiative transfer setup.
A paper that cites 15 of the draft's references, to cite where the radiative transfer is set up.

We checked a few of them against the reference list of the paper: it cites two papers by Izquierdo et al. (2021 and 2023), but not the 2022 one proposed here, nor Isella et al. (2016) or Hughes et al. (2017). The list is a set of candidates to judge, not a verdict: mark each one Added, Already cited or Not relevant. The exports leave out those marked already cited or not relevant, and a new search of the same draft leaves them out too.

4. Export, and search again before resubmitting

With the text kept, Search again before resubmitting: the papers this search had not found are marked New, and those you set aside stay aside.

Search again from the kept text of the draft, Delete the text of my draft, and Go further: watch this draft with Scoop watch, the bibliography check, a collection, the chat, a literature review and Who works on it.
Search again from the kept text, and the next tools.

Go further follows the draft until it is published: Scoop watch emails you the day a new paper comes close to it, and Who works on it? gives the teams on the question, and possible referees.

Limits

FAQ

Is my draft kept?
Only if you tick “Keep the text of my draft”: it is then stored encrypted, to search again later, and you can delete it at any time. Otherwise it is read for the searches and not stored. The list of papers found stays in your account.
My bibliography style prints no titles. Does it work?
Yes. A reference is recognised by its arXiv id, its DOI or its title, and, when the style prints no title (APS, AAS, MNRAS…), by its first author and year: those are set apart as probably cited, never proposed. With your .bib, every reference is recognised exactly.
How does it find papers written with other words?
Besides searches written from your question, method and data, it adds the papers that cite several of your references (bibliographic coupling) and the founding papers the closest candidates cite most often.

Sources