Assembly challenge

Last year the assembly challenge was too easy, with too high quality data. As a result most assemblers gave very similar results, making it hard to have a "competition"

For this year we should find a dataset that is a bit more challenging (high repeat number, weird GC content?)