I have encountered a strange issue in the alignment of a whole genome from Illumina. The below image shows PE reads mapped using Eland. Blue indicates base quality scores between 0-9 and white >=30. My question is this: What could possibly cause the drastic drop in base quality scores toward the 3' end of these reads to align in such an uncanny way? Does this look like and indel or translocation, or is this just the typical 5'->3' quality drop you would expect?