Writing a paper

Many journals have specific requirements so it is a good idea to write with a specific journal in mind. For example, some journals require the methods section to go at the end, which will affect how you write the results.

Modelling papers can be particularly difficult as the traditional separation between methods and results is less clear. Stacey Smith? at the University of Ottawa has shared a presentation about how to write a modelling paper, with 13 stages of a draft paper to show its evolution.

Most papers we write will be a collaboration so don’t be afraid to ask your colleagues and co-authors for help.

Authorship

The International Committee of Medical Journal Editors has defined authorship as requiring four criteria:

• Substantial contributions to the conception or design of the work; or the acquisition, analysis, or interpretation of data for the work; AND
• Drafting the work or reviewing it critically for important intellectual content; AND
• Final approval of the version to be published; AND
• Agreement to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

Many journals subscribe to these and require that everyone who meets the criteria is listed as an author and no one who does not meet the criteria is listed.

Pre-prints

Most journals allow work to be shared as a pre-print without affecting eligibility for full publication in the journal. Many pre-print servers are available. medRxiv is the most relevant for most of our work. bioRxiv and arXiv may also be relevant.

Writing tips

• Present your work. Organising your thoughts for a presentation at the weekly group meeting can help to clarify the story you want to tell.
• It’s a good idea to get into the habit of writing regularly. This can be easier than doing all your writing in one go. For instance, you could set yourself a target to write for at least 30 mins every day.
• Topic sentences: the first sentence of each paragraph frames the topic of that paragraph. More here.
• Be receptive to feedback! Try to assimilate the big picture things that co-authors are getting at, not just the immediate message.
• Be specific - give numbers, examples, etc.
• Introduction
  • Inverted pyramid: start of with big picture, narrowing down to the specific issue/question you addressed in this paper
  • E.g. for my first draft I try and follow something like (though the final version may look substantially different, having a simple structure helps me get started):
    • Paragraph 1: X is an important issue, because A, B, and C
    • Paragraph 2: This has been investigated in a number of ways. Author D found such and such, while author E found blah blah.
    • Paragraph 3: The role of Z is not currently well understood but is important to consider because of …
    • Paragraph 4: Here we investigate Z. Specifically we … {give details}.
• Discussion: I usually start off with one paragraph on each of the following, then expand/contract as necessary in the redrafting phase and following feedback:
  • Big picture summary of results
  • Interpretation and implications of results
  • Comparison to other studies, incl. Implications in light of our study and others
  • Strengths and limitations
  • Conclusions and next steps / future work
• This thread might be useful: https://twitter.com/vineet_chopra/status/1537843032287870977 which is similar to the approach described above but with more details.

Software

Writing software

• Microsoft Word / Libreoffice Writer
• Google Docs. Less features than the others, but good for collaboration. A good approach can be to draft the manuscript in Google Docs, and then download it as a docx or odt and edit that file for publication.
• LaTeX / Overleaf. LaTeX is useful if you are using a lot of maths, or to improve the reproducibility of your document (e.g. easier to automatically update figures if they change. Overleaf is the collaborative version of LaTeX, and the Overleaf website(https://www.overleaf.com/) has lots of templates and guides to get you started with LaTeX. Some journals also provide LaTeX templates to use for submissions.

Reference managers

• Zotero. Works with: Google docs, Microsoft Word, Libreoffice Writer. You can also export your references to a BibTeX for use with LaTeX. Open-source, and free to use.
• Mendeley. Works with Microsoft Word and Libreoffice Writer. Owned by Elsevier. Free to use.
• EndNote. Works with Microsoft Word. Does not work with Linux. Free through University of Oxford IT services..
• Paperpile. Works with Google Docs.

Keeping track of literature

• Some good advice in here
• It is a good idea to get into a regular reading habit. For instance you might set yourself a target of reading one paper a day on average. This may be higher when you are drafting an introduction or a proposal etc.
• It can also be worth occasionally reading papers outside of your immediate research interests as often you come across approaches to transfer.
• Most reference managers allow you to attach notes to papers. This can make it easier to keep track of your notes on a particular paper, and make them more easily searchable.
• You can also use your reference manager to keep on top of the papers you plan to read, e.g. by tagging them unread. This can be more easy to manage (and easier on your RAM) than keeping a bazillion tabs open or lots of pdfs saved on your desktop.
• To find new papers, some of the main ways include:
• Pubmed alerts
  • When you do a search in pubmed, you can save this search to your account and then get daily or weekly updates when anything new that fits those search criteria is published. Just click the create alert button to do this.
  • It can be useful to use the keyword categories in pubmed, such as [Title], [Abstract], [Title/Abstract], [Author], etc. in combination with boolean operators (OR, AND, NOT). For example, this search will return any papers with “antimicrobial resistance” in the Title or Abstract and at least one of Lim C or Cooper BS as an author.
• Follow scientists on twitter. This can be the best way to find pre-prints
• Google scholar alerts for individual authors
• Once you get alerts, one approach is to first read the abstract (or twitter thread) then if you think you need to know more, save it to your reference manager and mark.

The Bodleian library offers support from dedicated librarians in the Medical sciences to guide you on a literature search, you can reach them at hcl-enquiries@bodleian.ox.ac.uk.