What are the best steps in terms of structuring, planning, your actual writing?
Everyone has their own favoured approach to writing, so there is no ‘one-size-fits-all’ answer that applies to everyone. That being said, there are some guidelines that will help scientists to write a good paper with relatively little stress. First and foremost is that good planning is really helpful, both to you as you write, and to your audience of editors and readers. If your writing is unplanned and reads more like a stream of consciousness, it will be almost impossible to do full justice to your work, and will come across as confusing and messy to those who read it (remember that your readers will be much less familiar with your project than you, hence they will need it to be effectively communicated to them).
In my opinion, the most useful first step is to formulate your introduction (either as polished text, or at least as bullet lists of all the key points you want to make in it). The introduction sets the scene by introducing the broad field, describes where your subfield stood in the run up to your project and what key gap or question was the burning issue that you set out to address. You’d then close out the introduction with a brief big picture overview of what you did and what you found (just a few sentences). The reasons I think that formulating the introduction is a great first step are that 1) it forces you to think logically about the big-picture aspects of your project and its motivation, which will help you to think more clearly when you do all the subsequent detailed writing; and 2) it serves as a useful conceptual anchor – keep it in mind throughout the rest of the paper and avoid going off on unnecessary or distracting tangents.
After formulating your introduction you’ll be in a good position to work through the results section. This should be detailed, but clear and understandable. Think logically how the results/conclusions of one experiment lead on to the next one, so that it is clear to readers why you did each experiment. Writing according to the most logical conceptual progression of experiments is probably more important than describing everything in strict chronological order. Materials and methods sections could be assembled around the same time as the results – ideally you will already have strong lab notes on the methodology that just need to be adapted. As you write the results section, keep an eye open to the materials and methods that were used, as this will make sure that you don’t miss any out in the materials and methods section.
The discussion section should be easy to formulate once the introduction and results sections are written. Given the state of the field described in your introduction, and now with the addition of your described results in the results section, where does the field now stand? Do your results help to settle a debate, confirm a prevailing model, or are there discrepancies with previous studies (and if so, are there any logical explanations for it)? What are the limitations of your study and/or alternative interpretations? What are the possible future directions of your work and the wider field?
Thanks to Darren for a really useful answer. For more information on writing the various sections of a scientific paper, as well as tips on structuring your work, visit the module 'From Introduction to Conclusions'.