Error in coverting SRA to fastq

[yixf]

Hi,
Who is familiar with NCBI SRA Toolkit?
I downloaded a SRA data(SRR015256.sra) form NCBI. I want to convert it from SRA format to fastq format. When I use SRA Toolkit (version 2.0.b5, CentOS Linux 64 bit architecture) on Ubuntu like this:

./fastq-dump -A SRR015256 -D ./SRR015256.sra

it works! But after several minutes, when converting is nearly finished, it print error, like this :

fastq-dump err: row not found while executing function whitin virtual database module - Failed SRR015256

Why it can run, but can't finish? The toolkit has bugs?
Does anybody know the error's meaning? How to solve it?
Thanks a lot!

PS:
I run the toolkit on Windows7, it has the same problem!

[EGrassi]

The usage seems right to me (I am using the same on lite.sra right now). Have you checked if the file has been corrupted during the download?

[yixf]

The usage seems right to me (I am using the same on lite.sra right now). Have you checked if the file has been corrupted during the download?

I don't have md5 to check out. So I don't know whether the file has been corrupted during the download. But I test several files, all them have the problem: can work, but can't finish!
BTW, I find fastq format files can be downloaded from NCBI SRA. Now I can keep away from the sra toolkit.
Thanks a lot, EGrassi!

[gabrielw]

Here isnt working, Ive got UBUNTU 9.10 and that happens

gwajnberg@bioinf05:~$ ./Downloads/sra_sdk-2.0.0b6/linux/rel/gcc/i386/bin/fastq-dump -A SRR002099 -D ./srr/SRR002099.sra
2010-12-13 16:54:06 ./Downloads/sra_sdk-2.0.0b6/linux/rel/gcc/i386/bin/fastq-dump err: param excessive while constructing memory map within file system module - failed to open './srr/SRR002099.sra'

[yixf]

Here isnt working, Ive got UBUNTU 9.10 and that happens

gwajnberg@bioinf05:~$ ./Downloads/sra_sdk-2.0.0b6/linux/rel/gcc/i386/bin/fastq-dump -A SRR002099 -D ./srr/SRR002099.sra
2010-12-13 16:54:06 ./Downloads/sra_sdk-2.0.0b6/linux/rel/gcc/i386/bin/fastq-dump err: param excessive while constructing memory map within file system module - failed to open './srr/SRR002099.sra'

I didn't meet this problem.
RAM is not enough?
PS: I run it on AMD64 Ubuntu 10.10.

[tgoldman]

Hello,

I am having trouble with one particular experiment (SRX017810) in converting the SRA read files (SRR038114, 5, 6, and 7, all about 500MB) to fastq. I haven't had any problems with other files converting, including some files over 3GB.
I am running RHEL5, 64-bit, 8GB RAM.
When I run fastq-dump, it appears nothing is happening, while other files you can see the resultant fastq file being created right away.
Also, is is still possible to download static fastq files from NCBI SRA? yixf posted: "
BTW, I find fastq format files can be downloaded from NCBI SRA"

Thanks. Any help would be appreciated!

[yixf]

Hi,tgoldman
You can download the fastq file by following steps:
1.Go the the SRA homepage, and enter the key word(here:SRX017810) to search.
2.The new page will show, and only one record is found.
3.On the right panel, there are four SRR* record, click one of them, for example, SRR038114.
4.Now, another new page will be shown to you.
5.On the middle, after "Find spots: ", there are two buttons, click the "Download" one.
6.Another new page ...
7.On the left panel, select the record you want to download; on the right panel, cancel the "filtered" check box and change the format to "FASTQ".
8.Now click the "save" button, you will get your data on fastq format.
9.After you get the file, please pay attention to this: the fastq file has a empty line at the tail(you can see this by "tail *.fastq"). I don't know does this affect next analysis, but I delete this empty line in my file.
Hope this will help you.
Gook luck!

[tgoldman]

Thanks yixf!

[yixf]

Thanks yixf!

It's my pleasure!

[yixf]

New version of SRA toolkit has been released(release candidate on December 29 2010, version 2.0.0rc1). The compiled binary can be founded at SRA home page.
I tested the "CentOS Linux 64 bit architecture" on "Ubuntu 10.10 desktop, AMD64" using the test file of "SRR015256.sra". It works! The "bug" has been fixed!
Thanks the developers!
BTW, I reported the "bug" to SRA toolkit's developers. Now, you can visit it at the link.I am sorry that I can't open the link so that I don't know its content. I don't know why I meet this problem.

[cascoamarillo]

Hi,
Please, let me continue with the post.
I have an error while dumping a SRA file. It's not the first time I use it, never having troubles before with other files:

$ fastq-dump --split-files -A SRR826660.sra
2014-08-06T21:42:59 fastq-dump.2.1.8 err: data bad version while constructing page map within virtual database module - failed SRR826660.sra
Written 0 spots total

It is not a memory problem.
Linux 2.6.32-358.11.1.el6.x86_64 (1TB)
fastq-dump.2.1.8

My best option, the file somehow is corrupted. The download seems legit. You can check it on

http://trace.ncbi.nlm.nih.gov/Traces/sra/?run=SRR826660

Any shot?

[GenoMax]

Have you tried providing full path for the file like so?

$ fastq-dump --split-files -A /path_to/SRR826660.sra

Working with sratoolkit v.2.3.2-4.

[cascoamarillo]

Updated to sratoolkit v.2.3.5-2....now it works.
Thanks.

[dena.dinesh]

I downlaoded the SRA files from the link
http://www.ncbi.nlm.nih.gov/geo/quer...i?acc=GSE57226 and when I tried to convert into fastq files I am always getting this error.

"fastq-dump.2.1.7 err: data bad version while constructing page map within
virtual database module - failed "

The command I used is fastq-dump SRAfile.sra

I use lasted version SRAtoolkit.

Very often I get this error for other datsaets also. how to solve it and is
it possible to download the fastq files from the ncbi or Arryexpress
database.

Kindly Guide me.

[GenoMax]

Have you tried providing full directory path for the .sra file?

$ fastq-dump /path_to/SRR1266967.sra