User Tools

Site Tools


wiki2016_rna_reproducible

Reproducible Research

Conduct reproducible projects …

  • Stay organized
  • Keep a lab notebook
  • Start good coding habits (commenting, testing, trapping errors)
  • Be VERY careful when opening anything using Excel or Word (autocorrect can mess things up).
  • Keep track of standard input files, software versions, package versions
  • Set seeds in random processes
  • Automate reports and log file generation
  • Avoid mouse-clicks, codify as much as possible
  • Use knitR with R projects

Publish reproducible research …

  • Publish Level I, II, III data, and metadata for each paper:
    • Metadata. spreadsheet-type information about the samples and project
    • Level I. fastq files
    • Level II. count matrices, normalized count matrices, .wig or .bw files, .bam files (if allowed),
    • Level III. processed results files (typically supplemental data tables)
  • Become familiar with NCBI GEO and SRA:
  • Become familiar with the standards of your field or journal
    • PLoS: Your high throughput sequencing data must be uploaded to NCBI GEO and SRA PRIOR TO submitting your manuscript for review
  • Write well organized scripts and make them available upon request
  • Upload scripts to a repository site (https://rpubs.com/, http://github.com)
  • Consider hosting a website for your published datasets
  • Learn to use knitr to produce documents for each time you execute your R code.
wiki2016_rna_reproducible.txt · Last modified: 2017/12/04 21:31 by erin