I have generated fastq files from Illumina MiSeq for bacterial genome sequencing. My reads are 2 X 300 bp paired end reads. I checked quality of the reads in FastQC and found that Q score goes below 26 for forward reads from base position 200 and Q score goes below 26 from base position 135 in reverse reads. Also k-mer tab gives failure (X mark) in FastQC report. Reads were around 4 millions for forward and reverse sequencing reads. My question what parameters do I need to consider for trimming poor quality reads (both forward and reverse)? Do I need to trim bases to same length in both forward and reverse reads? What are good parameters for improving QC before proceeding to de novo assembly.