We recently sequenced 96 samples on a Nextseq, prepare from single- and double stranded DNA, using the appropriate protocols.

Turns out that the samples using the single-stranded protocol seemingly did not work. One of the things I can see is that we got a very high frequency of long poly-G sequences (~20%). Manually checking the fastq files shows both complete 76bp poly-G sequences, and in other cases a small poly-A tail (10bp) following the indexing adapter, then followed by the long poly-G tail. Do these poly-G tails mean no signal (considering the NextSeqs two-colour chemistry) and should therefore be ignored and trimmed? Can they be the sympton of some library preparation problem (bad adapter fill-in, blunt-end repair)? Thanks in advance!