Improving the interface of our reproducible research reports

Another published research of the Lab goes fully reproducible with the manuscript full text, raw data and the R codes made available. Since the first reproducible report I made using R Markdown and deployed later last year, I was able to improve the interface of the HTML report which is now multiple-page.

This new report of a meta-analysis study uses R Markdown websites to produce a multiple-page HTML report. There are four main pages that describe:

  1. Pre-print version of the manuscript
  2. Raw data in csv format
  3. Commented R scripts
  4. Scripts for plots

Anyone may download files or clone the repository and reproduce the analysis, given the right packages are installed.

This MSc research published in Plant Disease also used meta-analysis - preceded by a systematic review to identify and select primary studies - to summarize the effects of specific fungicide (two triazoles and one benzimidazole) for controlling Fusarium head blight, a fungal disease that damages the grain and produces mycotoxins that are harmful to consumers.

The new report is intended, as before, to demonstrate all decisions taken during data analysis, from data import to transform, calculation of new variables, statistical modeling and preparation of high-quality plots for publication. It is also hoped that network meta-analysis, whenever multiple-treatments are under comparison, is better understood and used more extensively in the plant pathology field whenever appropriate.


comments powered by Disqus