I'm working with whole-genome sequencing data of a non-model organism. I ran the FastQC for the quality control of the data and I found faliures both in adapter content and tile's quality. Adapter content is not an issue anymore, but I'm struggling with per tile sequence quality. I tried to filter it with FastP and with filterbytile.sh script from BBMAP independently but both did not manage. I attach figures of one raw PE genome (with filterbytile script some of the red tiles disapear but still fails):
Reverse: picture 2
I know there are not a lot of tiles with red colours, but I just wanted to be sure if I should try to filter them or keep them like this.