Finding enriched Gene Ontology (GO) terms or pathways from a gene list.
Enrichment Analysis can work on Gene Ontology (GO) categories, canonical pathways, cytobands, miRNAs' target genes, transcription factors' target genes, and so on. David Functional Annotations is a free web tool and we recommend you use it, and Subio Platform plug-ins have some more useful features.
The result table has Redanduncy90 score, which indicates how much redundant the measurement list is. It is the 90th percentile of the number of overlapping measurements against other searching lists. If you want a set of measurement lists which are relatively unique, select lists with the lower Redanduncy90.
The Limination of Enrichment Analysis
Enrichment analysis is a method that relies on a database of curated information from papers. In other words, therein lies the fundamental limitation.
- Not all biological information is available to humankind. Enrichment analysis will produce biased results toward fields that have been heavily funded and heavily researched in the past. Veiled knowledge is never come up by enrichment analysis.
- Not all of the information found in the analysis are documented. Only a tiny portion of the information is mentioned in papers. And the selection of genes to be written is picked subjectivity under the influential paradigm.
- As omics become more widespread, many papers with very dubious data analysis are being produced, so the reliability of the literature information itself is decreasing.
Many papers have been being produced that somehow do enrichment analysis and then list the relevant GOs and pathways to make it look good. However, enrichment analysis does not lead to any conclusion. Instead, it presents you hints to interpret the experimental results. Thus, it’s you who think about what and how to demonstrate the hypothesis brought by the enrichment analysis.