10mistakes

10 common mistakes that could ruin your enrichment analysis:

1. Using uncorrected p-values
2. Not using a background list
3. Using a tool that doesn't report enrichment scores
4. Prioritising results solely by p-values
5. Using gene lists that are too large or too small
6. Combining up and down-regulated genes in the same ORA test
7. Using shallow gene annotations
8. Using outdated gene identifiers and gene sets
9. Bad presentation
10. Neglecting methods reproducibility

Contents

• 10mistakes.bib: Bibliography

• Dockerfile: Dockerfile

• analysis.Rmd: Main analysis script

• f1000research.csl: Citation style file

• manuscript.Rmd: Manuscript file

How to reproduce

Start a container

docker run -it mziemann/10mistakes:latest

Once in the container, pull the latest code.

git stash && git pull

Now the two R scripts can be run to conduct the analysis and generate the manuscript.

Rscript -e 'rmarkdown::render("analysis.Rmd")'

Rscript -e 'rmarkdown::render("manuscript.Rmd")'

Copy the data back to the host.

docker cp $(docker ps -alq):/10mistakes docker_data

Now the html files and figures can be inspected in the docker_data folder.