Extracting reads from SRA file: constant quality scores?

[liz_is]

I have downloaded an SRA file and extracted the reads using fastq-dump, but in the resulting file, every read has the same quality score, which is constant along the read.

The file in question is this one: http://sra.dnanexus.com/runs/SRR489720

and the head of the Fastq file is as follows:

@SRR489720.sra.1.1 1 length=36
GCTTAATCTATTTTGCAGATGTAATTGTATGTATTA
+SRR489720.sra.1.1 1 length=36
IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII
@SRR489720.sra.1.2 2 length=36
CTTGTATGTACTTAAAATATTTTATCATAGTTTTCA
+SRR489720.sra.1.2 2 length=36
IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII


The read identifiers look wrong here too! It may be worth noting that I got an error when initially trying to use fastq-dump to extract the reads: 'This operation requires access to external reference sequence(s) that could not be located'.

I followed the instructions here (http://www.ncbi.nlm.nih.gov/Traces/s...lkit_doc&f=std) on running configuration-assistant.perl, and was then able to extract reads - but there's clearly still a problem somewhere...

[maubp]

At face value that is a really good score, PHRED 40 (the ASCII value of I is 73, assuming the standard Sanger FASTQ offset of 32, gives 73 - 33 = 40).

[liz_is]

FastQC detects the encoding is Illumina 1.5, in which I gives a quality score of 9. But with only one character present, that's not a reliable detection. The other datasets I've downloaded from this group use Sanger / Illumina 1.9 encoding, so assuming this one is the same, and given that scores > 40 don't really appear - it looks like all positions have been arbitrarily given the highest possible value.

Maybe it's a FASTA file, not FASTQ, that was uploaded by mistake or mislabelled...

[migs54]

I have downloaded an SRA file and extracted the reads using fastq-dump, but in the resulting file, every read has the same quality score, which is constant along the read.

The file in question is this one: http://sra.dnanexus.com/runs/SRR489720

and the head of the Fastq file is as follows:

@SRR489720.sra.1.1 1 length=36
GCTTAATCTATTTTGCAGATGTAATTGTATGTATTA
+SRR489720.sra.1.1 1 length=36
IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII
@SRR489720.sra.1.2 2 length=36
CTTGTATGTACTTAAAATATTTTATCATAGTTTTCA
+SRR489720.sra.1.2 2 length=36
IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII


The read identifiers look wrong here too! It may be worth noting that I got an error when initially trying to use fastq-dump to extract the reads: 'This operation requires access to external reference sequence(s) that could not be located'.

I followed the instructions here (http://www.ncbi.nlm.nih.gov/Traces/s...lkit_doc&f=std) on running configuration-assistant.perl, and was then able to extract reads - but there's clearly still a problem somewhere...

Did you resolve this matter. I have the same issue.
I directly downloaded the Fastq file a ChIP-seq data set using DRA search. I've been trying to convert/determine quality with no luck. When I run FASTQC the per base sequence quality and per sequence quality scores show no values.

Fastq:
@SRR039851.1 VAB_ChIP_H3K27me3_un_b11_76_396_F3 length=35
T01100020300101300200223312213020031
+SRR039851.1 VAB_ChIP_H3K27me3_un_b11_76_396_F3 length=35
!"""""""""""""""""""""""""""""""""""

I am currently using Galaxy to Compute quality statistics for SOLiD data after conversion of the data type to qualsolid. I want to map this data using bowtie.

My biggest concern however, is the quality score being constant across every base for each read.

Any suggestions?