I'm new in NGS and i'm currently working on RNA-seq. I had 6G of illumina reads and proceed with the data processing. After joining my paired end reads into a single file, I ran FastQC to get the median quality score before trimming the reads. But i found out the scores of my reads are very bad. I would like to know, should I continue to trim my reads by using 2 as the median quality score or map my reads without trimming them? the following image is the box plot for the quality scores across all bases of my data.
Any constructive advice/comments are highly appreciated. Thanks.