I used Trim Galore! on my GBS data and have ~30 overrepresented sequences in the fastqc report. Each sequence starts with the restriction enzyme's cutting site (PstI; TGCA). No sources have been identified and the percentages range from 10-77% for the R1 and R2 outputs. Is this to be expected or something I should be concerned about? Thanks.

UPDATE: I just realized that these percentages are out of 100, not 1, so 0.77 is less than 1%. That makes these results seem more normal.