Our genomics workflows use dozens of packages even before I get the data and start really doing analysis, statistics, and plots. It's just not feasible to cite every bit of code we use (though we certainly point people to the higher level routines, which they can use to see what was run/how to reproduce the results).