unknown bases in a RNA-Seq data

[arrchi]

Hi there,

I just got a RNA-Seq 50bp paired-end data generated by Illumina HiSeq. A strange thing is all the last base of the paired reads are N. For example:

$ head -4 m1.R1.fastq
@HISEQ:54:C1YCKACXX:1:1210:8881:74275/1
TACCTGGTTGATTCTGCCAGTAGTCATATGCTTGTCTCAAAGATTAAGCC
+
CCCFFFFDHHHHGJJJJJJJIJJHIJJJJIJJIIIJJJJJJIIJGIIIJI

$ head -48 m1R2.fastq
@HISEQ:54:C1YCKACXX:1:1210:8881:74275/2
AATAAATACACCCCTTCCAGAAGTCGGGGCTTGAATGCATGTATTAGCTN
+
CCCFFFFFHHHHHJJJJJJJJJJHHIJJJJJJIJJIJIJIIGGHJIIIE#
...
CCCCTTCGCGGGGGTCAGCGCCCGTCGGCATGTATTAGCTCTAGAATTAN
...
ATGAGCCATTCGCAGTTTCACAGTACATAGTTGCTTATACTTAGACATGN

I wonder what causes these unknown Ns? And why they only happen in one read of a pair-end reads? Is it worth trimming the Ns? If I don't, what will be the possible side-effects?

Thanks for any suggestions.

[mastal]

unknown bases in a RNA-Seq data

Hi,

It looks like something unusual happened during the last sequencing cycle if all of your bases are Ns.

However, with Illumina reads, base quality generally falls off towards the end of the read, and more so in the second read of a pair.

Have you looked at your data with FastQC?

You could try using trimmomatic to remove the N bases from the ends before
aligning the reads, but if it's just the one base at the end it shoudln't make that much difference.

Best wishes,
Maria

[swbarnes2]

It could be an instrument problem during the very last cycle.

Leaving the N in won't be a big deal.

[arrchi]

Thanks a lot for all your suggestions. FastQC per base sequence quality plot shows the last base of second reads is missing. I agree, the Ns only happen at the end of the second read, it won't make a difference if I l remove them or not.