Has anyone here looked at data from a paired-end multiplex run that used custom indexed adpaters (with the index between a standard adapter and the fragment)? I recently had a center run a HiSeq lane of 2 X 100 with 96 mutliplexed samples and am seeing a large proportion of fragments where the ends match different indexes, about 40%. I pooled my samples after ligation and finished the library prep using the pool, so my best guess is that there was a substantial amount of PCR-mediated recombination during the enrichment step.
Anyone else seeing this? Ideas on what causes it? Suggestions on how to avoid it in future runs?
Anyone else seeing this? Ideas on what causes it? Suggestions on how to avoid it in future runs?
