Once a draft of a genome sequence is prepared, a lot of bioinformatics work goes into finding the genes and other functional features or important loci that are in a genome. These annotations are numerous, difficult to perform and verify, typically take lots of expertise and time, and are not something we would want to repeat. So, genome project consortia will typically share their annotations in some way, often through public databases of some sort. BioMart is a common data structure and API through which annotation data is made available. In this recipe, we'll look at how to programmatically access such databases so we can get annotations for genes that we are interested in.