Give GO annotation a go - Part I

It was great to see so many participants at the functional annotation workshop at PAGXXVII (https://plan.core-apps.com/pag_2019/event/d42319e9aae3945c459f3e5e1526585a). Interestingly, many of you came to us expecting to learn how to mass annotate RNAseq data with Gene Ontology (GO) terms. I know the feeling: you have completed your awesome RNAseq experiment and have a long list of genes to analyze. You might be looking to see what GO terms are enriched or just want to know if your top candidates share some functions in common. Maybe you have noticed that sometimes your top candidates have only limited associated GO annotations and you might be wondering why this is the case in this age of large data.

The answer is simple: while some functional annotation can be done using computational methods, the linking of gene products and GO terms based on experimental information is still done manually. At TAIR and other biological databases engaged in functional annotation based on the peer-reviewed literature, functional annotation is detail-driven and laborious process where curators actually read papers to identify the experimental results and the appropriate GO terms for the genes that are described.

In future blog posts we will go into more detail regarding what it takes to curate a paper and assign a GO term including what you can annotate with GO, GO phenotypes, and choosing the right GO terms.