I need to have a RNA-Seq dataset and therefore, I've visited the following site NCBI-geo C. Elegans

In the Supplemantary file part, I clicked the SRP/SRP051/SRP051702 ftp and downloaded sra file. Then I need to convert it to fastq file format. For this purpose I've heard that there is a sra-toolkit and within it there is fastq-dump exacutable. However in order to use it, first of all I have to figure out whether my sra file contains paired-end or single-end read data. So my question is that how could I know the type of reads in the dataset (in the first link) ?

Without any information, I used --split_files flag of fastq-dump exacutable and it generated two 14 Gb (both of them are exactly 14.346.367.840 bytes) files and their name are SRR1741330_1.fastq and SRR1741330_2.fastq Does it mean my dataset is paired end ?

As an another question (different but related with part 1). In the SRR1741330_1.fastq file, at some lines, sequences contains different characters such as CCCFFFFFGFHHHGJJJJI#1?FEIGGI... Before looking into these files, I have thought that these sequence lines should only contain the letters of A,G,T and C . What are these F,H,J,I,#,? etc..

My questions might be trivial and unmeaningful but since I'm a totally new person in this area, I could not understand them.


New question

I wonder one more thing, since its related with previous part of my question I asked it here instead of creating new question. After I generate fastq file with dump_fastq executable, all the sequences are length of 50. Does it because of parameters of the dump_fastq executable or its related with original sra file? In other words, could I increase this length?


1 Answer 1


To find out whether a dataset was paired-end or single-end, go to SRA, click on a run, and look under "Library". Paired-end datasets will typically have "Layout: paired". Note that people don't always mark this correctly, which causes no end of headaches.

Regarding lines like "CCCFFFFFGFHHHGJJJJI#1?FEIGGI", that's the quality score line. Look at the fastq article in wikipedia for more details.

Edit: Regarding the updated portion of your question, no you can not increase this length. The sequences produced by Illumina machines are a fixed length and that length (or whatever length was submitted) is what you're getting. As a general rule, you'll get whatever sequence was updated (unless you forget the --split3 (or whatever it's called) option).

  • $\begingroup$ @Devon_Ryan thanks for your help, by mean go to SRA, I think you refer to sra file that I downloaded from the link I provide above,right? $\endgroup$ May 8, 2015 at 17:16
  • $\begingroup$ Okey, now I found it, it is in the browser $\endgroup$ May 8, 2015 at 17:33
  • 1
    $\begingroup$ Exactly, I mean the webpage for a run on the SRA website. The same information is typically available in GEO too (click on one of the samples and read the "Data Processing" section), but it's typically a bit faster to just go to SRA. $\endgroup$
    – Devon Ryan
    May 8, 2015 at 17:54
  • $\begingroup$ @Devon_Ryan thanks again, could you look at my question again, I've updated it $\endgroup$ May 8, 2015 at 17:57
  • 1
    $\begingroup$ I've updated my reply. The short answer is that you're getting what was uploaded, which is typically all that was sequenced. $\endgroup$
    – Devon Ryan
    May 8, 2015 at 18:30

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .