"A meek endeavor to the triumph" by Sampath Jayarathna

Wednesday, April 23, 2014

Quick and easy way to write a research paper publication

First prepare the figures.
Be sure that everything has been done n=3 or more, and be sure to do the statistical analysis (t-tests to compare 2 things that have roughly Gaussian distributions, Mann-Whitney otherwise; use ANOVA if there are three or more things you’re comparing). Something is different if p < 0.05, otherwise its not a statistically significant difference. And remember to put size bars on all the micrographs.


Then sit down with each figure and write a results paragraph for it
Results section – you can write a new paragraph every 5 minutes.  If you have writer’s block, sit down with the figures, and have someone take dictation as you answer the following questions for each figure.
The following set of sentences allows one to easily create a typical results paragraph:

  • what we wanted to find out
    “To determine if extracellular cAMP regulates.......”
  • what has been done before
    “We previously observed that ...”  (these first 2 sentences can be swapped) 
  • how we did the experiment (don’t give details, as these are in the Methods)
    “Control and xyzA cells were starved for six hours and cAMP was added....” 
  • what we saw, starting with a connection between the control and what has been done previously (if applicable)
    “As previously observed, wild-type cells showed a significant decrease in …” 
  • and then compare the mutant or treatment to the control
    “Unlike control cells, the xyzA cells showed no cAMP-induced....”
    or
    “Compared to the control, cAMP caused a significant increase in …” 
  • what this means
    “The data suggest that XyzA is required for ....”
Congrats, you’ve done a paragraph for your first figure!  Now repeat for all the other figures.

Then for each figure do a methods section. 
Pretend you are writing a lab protocol for someone – list exact composition/ concentration, volume, time, temperature if not room temp to do that experiment.  If there is a published paper that has the procedure in it, or you did something following the manufacturer’s directions, just reference the paper or write “..was done following the manufacturer’s directions.” 

The introduction will now be easy to write.
Print out the Results and Methods sections.  Get a red pen and circle all the big words or unusual techniques, and write a simple explanation for someone not in the field.
So for a paper showing that filamin is required for normal circadian rhythms in Dictyostelium, I’d write 3 paragraphs, adding references for pretty much each sentence or statement
  • what’s known about Dictyostelium
  • what’s known about circadian rhythms, and what’s known about them in Dicty
  • what’s known about filamin in general, and what’s known about its function in circadian rhythms and in Dicty
Depending on how things sound, you can swap the paragraphs around; start with the broadest/ most important topic and narrow things down.  So maybe you could start with circadian rhythms, then Dicty, then circad rhythms in dicty, and then filamin.

The Discussion is also easy.
Gather the ‘What this means sentences from the Results and glop them together for the first Discussion paragraph. 
  • “In this report, we have shown that ……”
  • Now start going through that first paragraph of the Discussion, connecting each of the sentences to other things in the literature
At some point do the figure legends, and you’re pretty much done!

This is an extract from Professor Richard Gomer's how to write a paper. 

No comments: