why all the reads in fastq files have the same quality value?

Question

[bright602]

Hi there, could any one tell me why all the reads in my fastq file have the same or very similar quality value as showing below? Thanks a lot.

@NB502075:3:H2FM3BGX3:1:11101:9808:6370 2:N:0:TAGAACAC TTCGCCTAGTGGGAGCTTATCTTCTCGCCACCCTCGGTGGAAATGCCTCCCCATCTGCACAAGATGTCTTTAAGG + AAAAAEEE/EEEEEEEEEEEEEEEEEEEEAEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE @NB502075:3:H2FM3BGX3:1:11101:7933:6371 2:N:0:TAGAACAC TCAGGATTCTTTGAGTCTTGGATAATGGCTCCCGCTCGTAAGGGAAAGGCTAAGGAGGAACAGGCTGTCGTGTCC + AAAAAEEEEEEEEEEEEEEEEEEEEEAEEEEAEEEEEEEEEEEEEEEEEEEEEEEEEEEEAEEEEEEEEEEEEEE @NB502075:3:H2FM3BGX3:1:11101:19490:6372 2:N:0:TAGAACAC TCGTAGGTGACGGGTTTTTCCTATGCCGAAAGGTATCGGTAAACCGTTGAAATTCTTCCATGTCCGGGATAGGGT + AAAAAEEE/EEEEEEEEEEEEEEEAEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE @NB502075:3:H2FM3BGX3:1:11101:3601:6372 2:N:0:TAGAACAC ACGTCAACTCTACGCTGGCGTAAATCAATGTTGGAAACACTCGAAGAACAACGAAAAGCCGTCGAAGGTGAAGCA + AAAAAEEE6EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE @NB502075:3:H2FM3BGX3:1:11101:16612:6373 2:N:0:TAGAACAC AGTGCTGTTACCATTAGGACTGACAGATTGAAAGCTCTTTCTCGATTTGGTGGTTGGTGGTGCATGGCCGTTCTT + AAAAAEEE6EEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEEE

[Joe]

Where did they come from? Straight off the machine?

[bright602]

Follow up the original question, when I ran through fastqc, there is no error bar in the per base sequence quality. So the graph looks not usual.

[Joe]

Plase use ADD COMMENT instead of answer unless you are actually answering your own post.

[WouterDeCoster]

Indeed. I have now moved his answer to a comment.

[Joe]

Can you provide a bit more information about the sequences? What sort of machine? What library size/type of sequencing is it? Does all of your data look like this?

[Devon Ryan]

The machine ID is for a NextSeq ("NBXXXXXX").

Answer 1

[Devon Ryan]

Since this is a NextSeq machine, please note that base qualities are binned into a few discrete values in that machine. For that reason, as long as the library is reasonably good quality, you'll relatively rarely see the error bars in FastQC reach much past 1 bin (a change of 5 on the Phred scale). For comparison, I just pulled a random sample from a recent NextSeq run and see 0 variance for the first ~30 or 75 bases in FastQC. I only see a yellow box in the last base. If you have a really messy library then you'll get more variance in that plot.