Creating scientific figures is essential for effectively communicating scientific concepts and research findings. However, it can be especially challenging, as it involves an inherently creative and abstract process.
Here, I would like to provide an overview of simple steps to creating a scientific figure (Warning: This doesn’t aim to be a data visualization post, but rather a general guide for illustrative figures, graphical abstracts, etc).
Define the Purpose: do you want to illustrate a process, summarize data? What is the main takeaway you want your audience to have from your figure?
Get Inspiration: have you seen other figures that you like that accomplish a similar purpose? Gather them and take note of everything that helps them to effectively communicate a concept
Gather Data: Collect the information you want to visualize in your figure. This could include biological processes, experimental methods, results, statistical analysis, or conceptual relationships, etc
Make a Sketch: In paper, make a rough sketch with the layout, elements, and overall composition of the figure (Extra tip: if you are specially bad at drawing like myself, you can also use: https://www.autodraw.com/)
Choose a Tool: Select a tool or combination of tools that suit your needs. Some options are:
BioRender provides both editing tools and several scientific vectors
Sketch Your Draft: Try to copy your first sketch and move elements to ensure hierarchy of information, avoid clout, etc.
Add Details: include clear labels and annotations to help viewers understand the content of the figure.
Color and Style: use consistent color schemes and ensure readability. Also, you can check your palette is colorblind safe here
Review and Refine: step back and take a look at your figure. Is it conveying the intended message? If possible, show to a friend or colleague. Make any necessary adjustments
Export and Format: export your figure in a suitable format and make sure it fits within publication guidelines (e.g., resolution, size).
Cite Sources: if you used existing content provide proper citations.
Some extra resources:
Here is a very nice guides to create figures from Nature and also, some examples of good data visualization/figures here.
Hey @DCoP_Innovators what tools have you used recently to make your figures? Is there an example of a figure you saw that made you say “Now, this is a good figure ”? Share!
This is a content of EXCELLENT quality. I’m saving all your references and step-by-step guide for further use! Thank you a lot for this!
Regarding your question, I mostly create my figures using Python and R, directly from data. I know that Adobe Illustrator or similar software are necessary to improve those figures, however, I’ve never dealt with them before. I often find myself also asking ChatGPT for suggestions on figures and helping me in my code to make the figure the best as possible.
I have two figures to share. The first one relates to the distribution between age of symptom onset from different movement disorders syndromes that consulted in our hospital and years of delay in order to obtain a proper diagnosis. The second one is from an article that’s currently in preprint and under review where we analysed the prevalence of PD in low to middle-upper income countries (https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4662346). Both are bubble plots.
This figure shows that patients with rarer movement disorders syndromes (such as myoclonus and paroxysmal dyskinesias) face a significant diagnostic delay in comparison to other syndromes, and that dystonia is the syndrome that a proper diagnosis is achieved much faster than others.
This figure highlights how the prevalence of PD and countries GDP per capita are correlated (even though also with a participation of other variables not here described). Studies from specific globe regions are highlighted in a different collor pallete.
@paularp excellent post; thank you! I really enjoy data visualization and put a lot of effort into my figures so that they are easily digested by the audience.
What got me started in data visualization practices was the works of Edward Tufte. I highly recommend his books! They have great recommendations for use of color, layout, typography, etc.
I am delighted about this post as an creative who is interested in science! Scientific illustrations and graphic design are my passion. I got my first honourable mention in a review paper I made illustrations for while in final year undergraduate (The therapeutic landscape of HIV-1 via genome editing | AIDS Research and Therapy | Full Text). I went on to make illustrations for an immunology book. I have developed this skill and look forward to opportunities to exploit. Having a biochemistry degree and almost a molecular medicine masters makes it easy for me to understand the scientific principles and my graphic skill allow me to express myself. I use PowerPoint for simple designs and Adobe Photoshop . I am happy to see that someone recognizes how useful PowerPoint can be for designs.
In terms of my workflow for creating figures and diagrams, I use a blend of R/ggplot2 and a Mac-only program called Omnigraffle (I’m not positive, but you may be able to substitute an Adobe product for Omnigraffle).
For data figures, I plot them using the ggplot2 package and get them very close to the end product in mind. I then will save out the figures in *.svg format. Omnigraffle allows for the import of *.svg files. The advantage of importing the *.svg figures into Omnigraffle is that each component of the figure is then modifiable, allowing you to change the text, size, color, shape, line thickness, etc. of anything and everything. I’ve noticed that this workflow from R → Omnigraffle allows you to incorporate the feedback from collaborators quickly without having to go back and forth from programming to editing.
For diagrams that do not require data, I make these in Omnigraffle (e.g., protocol/CONSORT diagrams).
I’ve posted on social media a few times about my love for Omnigraffle that it actually landed me a brief podcast interview about the product and how I use it for scientific publishing (feel free to listen here if you are interested: link).
Great and very interesting topic.
I think that for me is very important to understand things visually.
I use Biorender. and try to talk through the images. This week I had to summarize in 45 minutes, the progress I did during my 6 month stay at NIH, so I created a figure for the outline, and then I copied part of the figure in each corresponding slide, so it would make easier to the audience to follow a presentation, with so much information, and also very different. Here it is:
”? Share!
