When sequencers are processing sequence fragments, if there is little diversity at a particular bp locus this can cause problems for the sequencer. I have the superficial understanding that when a flow cell contains a significant percentage of a single base on any given cycle, it has difficulty deciding what base is present in each sequence. Why is this?
I think the problem is more that the software has a hard time identifying individual clusters if way more than a quarter of them are lighting up for a given base. If you have more diversity in the first few cycles, when the cluster coordinates are being identified, I think this is less of a problem.