Problem with bad quality bases in RNAseq

Question

[silas008]

Hi

I did a RNAseq to small RNAs (18-30nt). But the last (27-30) bases are with bad quality. Should I to cut that reads or not? I trimmed the last reads to obtain 18-25 reads and analyzed with this approach but I don't know this is correct.

[Istvan Albert]

What is your read length? If it is long then you would cut the ends anyway.

[silas008]

Reads = 36nt. Shoul I cut the ends to obtain reads 18-30nt?

[Istvan Albert]

Well if your reads are 36nt and 30 bases out of that are bad it will be a challenge to get useful information out regardless what you do.

[h.mon]

I think he means base 27 onwards are bad quality. And yes, if they are bad quality you should trim, but if base 26 is good why did you trim it?

[Istvan Albert]

I see, I misread that section. There is no problem them ;-)

[silas008]

Sorry. The bad quality scores are about 27 base, not exactly 27. But before 25 base I have very good quality scores.

Answer 1

[mark.ziemann]

Hi silas008, as a rule of thumb, you should be doing quality trimming (i.e., fastq_quality_trimmer) then adapter clipping (fastx_clipper) before mapping small RNA reads.

[silas008]

Thanks, Mark!

Actually, I did that before do the mapping of my data. But the fastqc keeping shows some problems. But now I think I understand the fastqc reports for DNA, not for RNA.