weird sequence representation from fastq-dump

[arielgis]

Hi
The output of fastq-dump looks like this (typical read)

@SRR415301.1 20110_run005Sample1_1_120_76 length=50
T3003332230231..31...022..20...02.0..0....0....0...
+SRR415301.1 20110_run005Sample1_1_120_76 length=50
!+%0(&2'&*+%&%!!/%!!!*%1!!%%!!!)%!-!!(!!!!)!!!!(!!!

When I use sam-dump I get something like this
AAATATCTAAGCANNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNNN

I find it hard to believe that the first is a digit representation of the former..

When I use bowtie I get an error:
Please truncate reads and quality values and and re-run Bowtie

Any help?

Thanks

[Richard Finney]

1) Looks like solid (color space) data. This may sometimes need to be converted. see solid2fastq.pl when use BWA for instance.
2) Please check the status of SRR415301 at http://www.ncbi.nlm.nih.gov/sra?term=SRR4153013 Is this record still valid?
3) When using the SRA tools, abi-dump is sometimes more usefull than fastq-dump.

[arielgis]

Thanks!

Hi

I will try using abi-dump and bwa instead of bowtie...

Thanks a lot

[carmeyeii]

Hi!

I am also trying to convert some Solid reads from .sra into regular fastq. Using fastq dump gives me the output I have pasted below. But when I try to run ABI-dump, all I get is:

carmen@genomics:/localdata2/carmen/RatAging$ abi-dump SRR388234.sra

2012-12-03T01:20:41 abi-dump.2.0.3 err: header bad version while validating header within database module - Failed to open index 'skey'

2012-12-03T01:20:41 abi-dump.2.0.3 fatal: SIGNAL - Segmentation fault

Aborted

Output from fastq-dump.

@SRR388232.1 F3 length=50
T30221030212303132201302210212.320221212302132.3113
+SRR388232.1 F3 length=50
!+)+-2)<)*;1?&%+<(0%(%)))8.(%1!=1%),&.-4/@.1%%!)/%7
@SRR388232.2 F3 length=50
T31000201030122203223112230122.232233222330122.3023
+SRR388232.2 F3 length=50
!'1(%&/+.%.)<,,%+)+(5-2*&/-,,&!4:1(,*'('%%**()!(()&
@SRR388232.3 F3 length=50
T32221211102122232312332123131.313223322313322.3312
+SRR388232.3 F3 length=50
!8'317/=0+-'2(%%%1-,+>@/%44%%%!?20(&<9'&.-@1//!@%3;

Does anybody know why abi-dump might not be working? And so can I use fastq-dump and then convert to fastq with the perl script?

Even the fastq-dump command returns an error, but when I look in the directory, it is in fact generating the files. This is what I get with the fast-q command:

fastq-dump SRR388232.sra

2012-12-03T01:18:43 fastq-dump.2.0.3 err: header bad version while validating header within database module - Failed to open index 'skey'

Thanks a lot.

Carmen

[carmeyeii]

Bumping up.

[Richard Finney]

You probably need to update your SRA toolkit.
It seems to work fine with abi-dump.2.1.12 (the version I used).
You are using version 2.0.3 (muy viejo ... very old).
The current version is 2.2.0.
There are binaries for SRA toolkit ... but I typically do the "download, unzip, untar, configure, make" route. You may wish to use the binaries for your system. If they don’t' work, you'll have to compile on your own.This is intimidating to the uninitiated; but it works.Check out :

http://www.ncbi.nlm.nih.gov/Traces/sra/sra.cgi?view=software

You can make and run your own, private version without "installing" it for everybody else. You'll need sysadmin (root) privileges to install for everybody.

________
my session :
wget ftp://ftp-trace.ncbi.nlm.nih.gov/sra.../SRR388234.sra

./sra_sdk-2.1.10/linux/pub/gcc/x86_64/bin/abi-dump.2.1.12 SRR388234.sra
Written 58777831 spots for SRR388234.sra
Written 58777831 spots total

-bash-3.00$ ls -tlrah SRR38823*
-rw-rw-r-- 1 finneyr finneyr 4.0G Dec 7 05:43 SRR388234.sra
-rw-rw-r-- 1 finneyr finneyr 4.9G Dec 7 06:16 SRR388234_F3.csfasta
-rw-rw-r-- 1 finneyr finneyr 9.3G Dec 7 06:16 SRR388234_F3_QV.qual
-rw-rw-r-- 1 finneyr finneyr 4.1G Dec 7 06:16 SRR388234_R3.csfasta
-rw-rw-r-- 1 finneyr finneyr 7.1G Dec 7 06:16 SRR388234_R3_QV.qual
____

Remember to use solid2fastq.pl if using BWA