December 2010 - galaxy-dev - lists.galaxyproject.org

Split FASTQ file into paired and unpaired reads
by Peter 15 Dec '10

15 Dec '10

Hi all, I've got a Python script which divides a FASTQ file containing a mixture of paired, unpaired and orphan reads into valid pairs and single/orphan reads. Such a situation can occur after applying quality filtering to a file of paired FASTQ reads. I've also had raw data supplied in this state from a sequencing center. It works by looking at the read names, and understand the /1 and /2 convention used by Illumina, .f and .r which I understand is common, and the Sanger convention as well. The only requirement (so that only a single pass though the input is needed) is that the input be sorted such that any pairs come as consecutive entries (forward then reverse read). Is there anything like this in Galaxy already that I have missed? Would you consider merging such a tool into Galaxy? I haven't written a wrapper XML file yet, and it also currently uses Biopython for FASTQ parsing, but I could switch it to use the Galaxy FASTQ code instead. Regards, Peter

2 3

Trackster BAM file Seg Fault
by Alan Jackson 15 Dec '10

15 Dec '10

Hi, I've been trying to use the Trackster visualisation with a BAM file, and find it segfaults quite regularly, with the (somewhat unhelpful) message ./run.sh: line 47: 13782 Segmentation fault (core dumped) python ./scripts/paster.py serve universe_wsgi.ini $@ Using gdb I have gotten the following info: Program received signal SIGSEGV, Segmentation fault. [Switching to Thread 0x4a6b5940 (LWP 13800)] 0x00002adeee8c7937 in get_chunk_coordinates (idx=0xffffffffb0018da0, tid=0, beg=0, end=-1341883296, cnt_off=0x2aaab0016600) at samtools/bam_index.c:501 501 samtools/bam_index.c: No such file or directory. in samtools/bam_index.c However the BAM index should be getting created and stored as part of the meta data. I also notice the pysam library being used is 0.1.1, whereas they are up to 0.3.1. Any help greatly appreciated, the visualisation looks very good when it's working. I am able to send the BAM file I'm using if necessary. Cheers, --Alan

2 1

Output files using format="input"
by Peter 15 Dec '10

15 Dec '10

Hi all, I'm trying to sort out some final details for my FASTQ filter by ID script before uploading it to the tool shed. This tool takes two input files, a FASTQ file (can be any type of FASTQ) and a tabular file (which has the IDs to filter on). The output file(s) are are also FASTQ (of the same variant as the input FASTQ file). I had simply used something like this when defining my outputs: <data name="output_pos" format="fastq" label="With matched ID" /> <data name="output_neg" format="fastq" label="Without matched ID" /> However, this discards the fact that the input (and therefore my output) may have been fastqsanger or fastqillumina etc. According to the wiki http://bitbucket.org/galaxy/galaxy-central/wiki/ToolConfigSyntax I can use format="input" in this setting... but here I have TWO input files (of different formats). Although this isn't giving an error, I find this results in my output being treated as tabular: <data name="output_pos" format="input" label="With matched ID" /> <data name="output_neg" format="input" label="Without matched ID" /> I've also tried using the names of my input parameters, but that is just treated as a new file format (interestingly no error is raised): <data name="output_pos" format="input_fastq" label="With matched ID" /> <data name="output_neg" format="input_fastq" label="Without matched ID" /> How is format="input" meant to work? Is it currently restricted to the special case of only one input file? Can you specify which input file somehow, e.g. via metadata_source? Thank you, Peter

1 1

32-bit python
by David Hoover 15 Dec '10

15 Dec '10

Are there any issues with using a 32-bit version of python for Galaxy? David Hoover Helix Systems Staff http://helix.nih.gov

3 3

FASTQ to Tabular tool doesn't accept plain FASTQ
by Peter 15 Dec '10

15 Dec '10

Hi all, I just uploaded a (Sanger) FASTQ file without specifying the format. Galaxy therefore treats it as "fastq". Some tools are happy to take this as input, but not the fastq_to_tabular converter. Please consider applying the following trivial patch: --- a/tools/fastq/fastq_to_tabular.xml Wed Nov 17 16:24:12 2010 -0500 +++ b/tools/fastq/fastq_to_tabular.xml Mon Dec 13 16:52:51 2010 +0000 @@ -2,7 +2,7 @@ <description>converter</description> <command interpreter="python">fastq_to_tabular.py '$input_file' '$output_file' '${input_file.extension[len( 'fastq' ):]}'</command> <inputs> - <param name="input_file" type="data" format="fastqsanger,fastqcssanger,fastqillumina,fastqsolexa" label="FASTQ file to convert" /> + <param name="input_file" type="data" format="fastq,fastqsanger,fastqcssanger,fastqillumina,fastqsolexa" label="FASTQ file to convert" /> </inputs> <outputs> <data name="output_file" format="tabular" /> Thanks, Peter

1 1

Invalid request ID
by nsfox＠uwaterloo.ca 14 Dec '10

14 Dec '10

When I try to import a sample from a csv file, it's failing and I'm getting the error message : Invalid request ID. Any suggestions as to what I might be doing wrong would be great. Thanks, Natalie

2 1

Nested structure or sections in drop down lists (select parameters)
by Peter 14 Dec '10

14 Dec '10

Hi all, I'm about to add all our existing BLAST nucleotide databases into the blastdb.loc file, plus all our protein databases into blastdb_p.loc, for use in Galaxy. I've realised these will result in some quite long drop down lists, which will not be very user-friendly. Currently in wwwblast, we have listed our BLAST databases by organism, and inserted dummy disabled options as spacers (e.g. species names). Perhaps a similar mechanism would work in Galaxy? According to the tool configuration page on the wiki the <option> tag only has value and selected tags, it would need something like an enabled tag (default true), and a way to select this from a loc file column. Alternatively I'm wondering if a (multiple select?) drill down select would be possible, something like that used in the Multiple Alignment Viewer tool (file visualization/GMAJ.xml). Are there any other similar situations elsewhere in Galaxy, and how was it handled? Thanks, Peter

1 0

Wiki erratum...
by Richard Bruskiewich 14 Dec '10

14 Dec '10

Galaxy Colleagues, I don't know who is maintaining the Galaxy wiki page at http://bitbucket.org/galaxy/galaxy-central/wiki/NGSLocalSetup but I noticed that the Python script under the Megablast instructions has an error: the "defline" operation after the "line.startswith" should be moved *after* the if length > 0 statement, otherwise, the defline is reset incorrectly before the previous sequence is written out. This results in a frameshift in the FASTA header line identifiers (i.e. the current sequence gets the next sequence identifier). I've commented out the erroneous defline below and added the right one: import sys length = 0 defline = '' seq = [] for line in sys.stdin : line = line.rstrip( '\r\n' ) if line.startswith( '>' ): # defline = line.split( "|" )[1] # defline should NOT be here if length > 0: print ">%s_%s" % ( defline, length ) print "\n".join( seq ) length = 0 seq = [] defline = line.split( "|" )[1] # defline should be here else: seq.append( line ) length += len( line ) print ">%s_%s" % ( defline, length ) print "\n".join( seq ) While on the topic of this page, perhaps the software versions need to be revisited. Megablast has been superseded already by Blast+. Perhaps new releases of Galaxy should update this? BTW, when is the new Galaxy release (cloud man AMI too...) coming out? I heard rumors that it was due this week. Cheers Richard Bruskiewich -- *Richard Bruskiewich, PhD* Senior Scientist, Computational and Systems Biology Applications Team for Computational Genomics T.T. Chang Genetic Resources Center International Rice Research Institute

4 3

Column order on Tool Shed
by Peter 13 Dec '10

13 Dec '10

Hi all, In Galaxy itself, a tool ID is shown (as a link) followed immediately by its description in the left hand column. As a result many tools have their description written as an extension of the ID text. For example, "Join two Queries" as the ID and "side by side on a specified field" as the description. In these cases the description on its own is not meaningful. This is a problem on the Galaxy Tool Shed since the column order is ID, type (tool vs suite), version, description, categories, uploaded by, and rating. I'd suggest putting the description column immediately after the name. Or maybe even combining the columns, with the ID in bold or something? Alternatively, perhaps Galaxy could try and use self standing descriptions? What do you think? Regards, Peter

2 2

Bug report for uploading tool updates to Tool Shed
by Peter 13 Dec '10

13 Dec '10

Hi all, I have an already approved tool on the Tool Shed, TMHMM and SignalP suite version 0.0.1 I uploaded a new tar ball for version 0.0.2, but did not click approve - I realised I needed to update the REAME file. So I deleted this update. [Note the UI could be clearer about what delete does, thankfully it did not delete the original approved version as well.] This aborted update now shows under "Browse your tools" with status "deleted". I now tried to upload a revised tar ball for version 0.0.2, but it complains "The new version (0.0.2) matches an old version. Check your version in the tool suite config file." The version it matches is the deleted one (my aborted 0.0.2 update). I would expect the version check to ignore deleted tools. Can this be changed? Thanks, Peter

2 5