Hello Chris,

Using a fasta file as input for LASTZ is the correct way to run the tool right now.

You discovered a small mismatch between our documentation and the version of the LASTZ wrapper on the main Galaxy instance. Fastq is not directly accepted (as the documentation states), so either using an original fasta file or using the "FASTQ to FASTA converter" tool would be required before running LASTZ.

Currently, LASTZ itself does not use quality scores for the alignment process, but will pass these values (ascii format) into the output file (SAM) for use in downstream analysis. The public Galaxy instance will likely be updated in the future to support this option, but there is no set time-line. To be clear, the alignment results themselves would be the same with or without the quality scores being passed through a fastq input/SAM output.

Our apologies for the confusion!

Best,

Jen
Galaxy team

On 9/26/11 7:49 PM, chris.how...@csiro.au wrote:
Dear Galaxy Team and users,

I have some 454 reads that I would like to map against a contig assembly
using LASTZ. I have already mapped the reads uploaded in fasta format
against the assembly but, as mapping the reads in fasta ignores the base
qualities that would be present in a fastq file, I am concerned that I
might need the base quality information that may be crucial in deciding
on ‘real’ SNPs later down the line. So, as LASTZ apparently recognises
fastq (*see below), I converted the reads to fastq using the ‘Combine
fasta and qual’ tool in Galaxy and I am now currently trying to map the
reads to the assembly. However, Galaxy would not recognise the fastq
reads in the LASTZ input page. So I tried to fool it by changing the
data type of the fastq to fasta using the ‘Edit attributes’ function of
the history. This kept the fastq info but allowed Galaxy to recognise
the file as input for LASTZ. However, this mapping has been running for
almost 24 hours now and so I am concerned that there is an error.

Is anyone able offer any help with why Galaxy does not recognise the
reads in fastq format prior to mapping with LASTZ?

**

Here are what the first two reads look like in fastq format:

@GIQ547K01A7QJK length=76 xy=0381_0142 region=1 run=R_2010_06_11_16_16_09_

GCTTCGTGTGCGACGACACTCGTCATCGACAACGCAAGACTGGCGCTATCGCAATTGGACACACAACATGTGACCG

+

27 19 17 17 18 19 11 14 14 17 19 17 22 17 17 14 14 14 17 19 17 19 17 17
19 19 25 25 22 17 12 13 14 19 21 21 21 27 21 19 19 17 17 19 17 24 25 22
20 22 22 17 16 16 12 12 12 12 19 22 17 17 17 20 20 22 27 21 25 22 20 20
22 21 16 12

@GIQ547K01AE4BG length=40 xy=0055_0266 region=1 run=R_2010_06_11_16_16_09_

GTGACTAGATACATGCAATCAATTGTCCATGTCATTCGAG

+

27 23 23 19 19 19 18 19 21 19 18 19 18 25 27 27 26 26 27 27 27 27 27 19
19 18 19 19 27 27 25 24 25 25 21 21 22 22 22 18

**

**

** Input formats (copied from the LASTZ input page in Galaxy)*

LASTZ accepts reference and reads in FASTA format. However, because
Galaxy supports implicit format conversion the tool will recognize fastq
and other method specific formats.

With thanks,

Chris

**



___________________________________________________________
The Galaxy User list should be used for the discussion of
Galaxy analysis and other features on the public server
at usegalaxy.org.  Please keep all replies on the list by
using "reply all" in your mail client.  For discussion of
local Galaxy instances and the Galaxy source code, please
use the Galaxy Development list:

   http://lists.bx.psu.edu/listinfo/galaxy-dev

To manage your subscriptions to this and other Galaxy lists,
please use the interface at:

   http://lists.bx.psu.edu/

--
Jennifer Jackson
http://usegalaxy.org
http://galaxyproject.org/Support
___________________________________________________________
The Galaxy User list should be used for the discussion of
Galaxy analysis and other features on the public server
at usegalaxy.org.  Please keep all replies on the list by
using "reply all" in your mail client.  For discussion of
local Galaxy instances and the Galaxy source code, please
use the Galaxy Development list:

 http://lists.bx.psu.edu/listinfo/galaxy-dev

To manage your subscriptions to this and other Galaxy lists,
please use the interface at:

 http://lists.bx.psu.edu/

Reply via email to