There is no rule that says that every FastQC category needs to be Green
before you can proceed with further analysis. FastQC results need to be taken in context of the experiment being done since the limits that are tested (which generate the pass/fail flags) are for normal genomic sequencing.
Please check our several blog posts from FastQC authors that should address most of your questions: https://sequencing.qcfail.com/software/fastqc/
Following two address:
Per base content: https://sequencing.qcfail.com/articles/positional-sequence-bias-in-random-primed-libraries/
Sequence duplication: https://sequencing.qcfail.com/articles/libraries-can-contain-technical-duplication/
Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
what to do with per base sequence content and sequence duplication levels?