So, you are writing a paper...

December 10, 2017

I remember when a wise Professor told me that, not so long ago, you would carefully make a clear draft of your scientific findings and   to a journal with a cover letter and "best regards"; and the editors would be very happy and after a few rounds of revision, your work would probably be published neatly.


Today, with the dissemination of pay-per-publish journals, renowned journals are being (correctly) much more critic. Please don't get me wrong. Critics were there since the beginning. The most pronounced difference is that nowadays; besides a researcher, you also have to enhance your literary and graphic designer skills. Of course, no journal expects a James Joyce or a Milton Glaser. However, understanding the proper way to expose your work (and eventually practicing a little bit on Inkscape) does not hurt.


From the wise editors of Nature Methods:

"In one definition, a scientific paper is a selection of data and its interpretation, represented as words and images, to deliver a scientific message(s) to a particular audience. You should ideally have an idea of the paper's key elements—the message, the audience, the data, inconsistencies in the data—even before you begin to write."


Inspired by that, I created my own personal "definitely don't do that list" below:


The Ten Commandments of Paper Writing:
1. Thou shalt not digress. Go straight to the point.
2. Thou shalt not write like I am writing this list. We know you are smart, you don't need to impress by using fancy words.
3. Honor thy collaborators. A good collaborative paper is always actually (surprise) done in collaboration.
4. Thou shalt not be over-technical in the main manuscript.
5. Thou shalt not attribute a "design by the creator". Seriously, "intelligent design" proponents will take over if nobody watches. They got tired of creating their own journals and are now submitting to Plos One:
6. Thou shalt not make identical figures. Just scroll a bit on PubMed, you will know what I am talking about.
7. Thou shalt not make figures with panels A-Z. The supplementary section exists for a reason.
8. Thou shalt not cherry-pick data. Be transparent!
9. Thou shalt make "corrigenda" and "errata" when necessary. Come on, everybody gets one or two lines of code wrong once in a while.
10. Honor thy reviewers. Of course, there are "those". But most of them are indeed helping you.


This post is not meant to give those standard set of rules like: "why the hell people still use 'aforementioned'... if it was mentioned, it was 'mentioned'". Instead, I will just describe the best and worst features in manuscript writing according to my point of view (it's my blog; I do whatever I want). Disappointed? Check out the links in the end of the post for a number of awesome straight-to-the-point tips on manuscript writing!


Best feature: Write for the world


Read commandment number two. Yes, that's exactly it. English is the global language, but not everyone is a native (I believe you can tell I am not just by reading this text). "Discombobulated", "Obnubilate", "Fulgurant", ..., please don't do that (unless the Nobel you are looking for is the literature one). Many non-Anglophones will read your paper (don't you want your ideas and findings to be disseminated the most?). Just as a reference:


Dark Blue: States and territories in which English is the first language of the majority of the population.

Light Blue: States and territories in which English is an official, but not the majority language. 


Worst feature: Over-technicality

Read commandment number four. My friend Dr. X (because I still want the friendship) is a genius. His MSc thesis had about 30 pages (note: in Brazil, it is common for a MSc and PhD thesis to have ~150 pages). Okay, his thesis was in the mathematics department; however, 20 pages contained basically matrices and Greek letters. Results? No. Why is that important? Figure out yourself. Again, don't do that please. Specially if you are in a multidisciplinary field such as computational biology. Unless it is a "novel computational/mathematical method's" paper, leave technicalities for the "methods" or "supplement" section. Most of your reader are interested in your results! So, do not waste people's time. Clear writing, clear figures and clear findings.


As a person who leans more towards the "computational" part, sometimes I do struggle to get some of the biology (although, nowadays not that much). But look at this piece of beautiful well-written paper by Schwarzer et. al.:



Clear, concise, and with clear biological terms. So, unless is a biological protocol or the details on your probabilistic approach to remove biases, go easy on technical terms.


So, what do you think? What are the worst/best features for you on a written manuscript? Any additions to the ten commandments? I'll add as much as you suggest.


Ow, and ALWAYS ask as many people as possible to look at your manuscript before submitting. This will certainly happen to you someday:



but it's for the best and will make reviewers and everyone happier.


Very good publication on "Elements of Style for Writing - Scientific Journal Articles" (thank you Dr. Argyris Papantonis!):


Great advice on writing a paper:
1. Perneger TV & Hudelson PM (2004). Int J Qual Health Care, 16:191-192.
2. Kallestinova ED (2011). Yale J Biol Med, 83:181–190.
3. Jha KN (2014). J Clin Diagn Res, 8:XG01–XG03.
4. Ashique KT & Kaliyadan F (2016). Indian J Dermatol, 61:26–31.


1. Editorial (2017). Nature Methods, 14: 1115.
3. Schwarzer W et. al. (2017). Nature, 551: 51-56.


Figure Sources:
Fig.4. Schwarzer W et. al. (2017). Nature, 551: 51-56.


Share on Facebook
Share on Twitter
Share on LinkedIn
Please reload

Please reload

Please reload


© 2017 by Eduardo Gade Gusmao