Assume you are writing a manuscript and you have uploaded your raw data to a relevant hosting site (e.g., The Dataverse Network, FigShare, etc.). Where in the manuscript should you indicate a link to this?
This question was posted on twitter and one suggestion was to include a link in the author notes and the results.
I am sure there are many ways to do this, but I have recently started to end the paper with a specific (unnumbered) section Further Material (or something like this). For an example, you can check this preprint of mine.
Before, I primarily gave a link to the online material as a footnote somewhere in the Experiments section, or wherever else it made sense from the text flow. However, it happened to me more than once that a careless reviewer overlooked these footnotes, and complained that the data has not been made public.
Some journals will also allow you to upload accompanying material to their website. However, at least in my field this is so unusual that I would be afraid that potential readers will not even look for the material on the publisher's website.
Some journals ask for a standardised data section - see, for example, the start of this recent PLOS paper. Some (eg JGR) recommend putting it in with acknowledgements. Many more (eg Marine Biology) are somewhat vague but suggest in-text citation of the dataset. It'd be worth checking what's standard for the journal(s) you're planning to submit to.
If there's no guidance at all, then putting in a short section at the end on accessing data is perfectly reasonable - it stands out, as xLeitix says, and it is easy enough to move it around elsewhere if requested. And if you have a DOI for the data citation, use it - it'll help credit the data if reused in future.
Edit: here's an excellent and detailed example of how to do in-text citation of data, from the American Meteorological Society. They explicitly deprecate the acknowledgements approach (though suggest both can be used in parallel if desired).
In epidemiology and other medical research it is extremely common to both describe the data set in question, and to discuss what software package was used to do the analysis, both of them in the Methods section of a paper.
As such, the most logical place is to link to open data is at the end of this description.
For example:
These data sets do not include patient-level information, but rather laboratory confirmed, suspected or probable cases of the disease, which is thought to represent the best available estimate of the current state of the epidemic. A curated version of this data is available at https://github.com/username/horribledisease.
