Splitting SQ runs so that they run through the pipeline better

[KimCallan]

Adam mentioned that when we ask for more than 1 lane of sequencing from the NovaSeq, it would be better that we pool our plates so that samples only run on 1 lane instead of over all the lanes requested. We can do this in the lab, but WALDO needs to be able to assign libraries on a plate differently then we are assuming so that the pipeline knows which group of libraries to assign to each lane