a primer in overrepresented sequences of RNAseq data

Question

[Takuma]

Hello

I got pair-end RNA sequence data in Novaseq 6000, using SMARTer Stranded RNA library by outsourcing. I checked the quality by fastQC, then found below overrepresented sequence in only reverse fastq.

---

GCCAGATCGGAAGAGCGTCGTGTAGGGAAAGAGTGTACGCGTTAGTGTAG

Illumina Single End PCR Primer 1 (97% over 34bp)

I'm wondering why it is single end PCR Primer, even though this RNA should have read by paired-end. I think the single end PCR primer comes from cDNA synthesis or library amplification, not sequencing.
Is it correct ?

And what dose 「97% over 34bp」mean?

Should I remove the primer to conduct de novo assemble by trinity. My purpose is detecting some genes of interests, not mRNA expression level.

I'm padawan learner of bioinfomatics. Regard