What is the minimum information to include in a functional analysis paper?

FAQ tags: 
Most journals require authors to submit high-throughput data to public repository as a pre-requisite for publication. As part of this process, the methods used to analyse data need to be reported in detail; this applies to both statistical and functional analysis. For papers describing enrichment analysis using GO, this means that the methods section should include the following information, to ensure the analysis is reproducible (an important criteria for reviewers' approval):
  1. What analysis tool was used and what version
  2. What statistical analysis method was applied, and what correction factors were applied if any
  3. Date or release version of both the GO ontology file and the GO annotation file used
  4. Background genome/proteome/dataset used in the analysis
  5. Whether any enriched terms were excluded from the results due to low numbers of query genes associated with the term (e.g., if you only included GO terms in the results which have more than 3 query genes)
  6. Please cite: The Gene Ontology Consortium. Gene ontology: tool for the unification of biology. Nat Genet. May 2000;25(1):25-9. Online at Nature Genetics http://www.nature.com/ng/journal/v25/n1/abs/ng0500_25.html
The supplemental data files should include:
  1. List of the IDs used, and also the IDs which were rejected by the analysis tool if any
  2. Full list of enriched terms
When undertaking the functional analysis and interpreting the results, consider:
  1. Is the number of genes analysed statistically valid? Or is the number too small to observe enrichment, or too large for the enrichment to be meaningful. (E.g., for a microarray experiment with ~25,000 interrogated transcripts, it would be difficult to observe enrichment when analysing less than 150 IDs; if the list were longer than 3,000 IDs, clearer results would require further filtering, e.g. based on significance threshold or fold change. This is only a rough guide.)
  2. Consider using more than one functional analysis tool, as well as fine-tuning the parameters used. You may also wish to look at overlap between results from different approaches.
  3. Think about the biology. E.g., if you need/wish to make a choice among enriched terms to show in a summary table, use descriptive GO terms. Do not only pick terms you're particularly interested in, and consider that very broad or generic terms, such as 'metabolic process', can be uninformative.