Galaxy-qld FAQ

This page has questions repeatedly asked by Galaxy-qld users.
You may also read GVL_FAQ page.

Q: Prokka jobs fail with the following error message:
Fatal error: Exit code 2 ()
Please rename your contigs OR try ‘–centre X –compliant’ to generate clean contig names.
A: Use shorter names, eg just one character, in Locus tag prefix (–locustag) and Sequencing Centre ID (–centre) settings.

Q: I cannot upload data to Galaxy-qld by sftp. What should I do?
A: Use ftp or ftp client. Windows File Explorer can be used as an ftp client. On command line type the following:
User name: email used for registration on the server
Password: your Galaxy-qld password
Use put command to upload a single file or mput for multiple files.

Q: Why output of pileup is not visible to other tools, such as SAMTools Filter pileup?
A: An old version of pileup assign tabular datatype to the output files. Change the datatype to pileup: click on pencil icon (Change attributes), switch to Datatype tab, and select pileup in the pull-down menu.

Q: How I can get news about Galaxy-qld or workshops run by GVL-Qld?
A: Follow us on Twitter: @GVL_QLD. You don’t need to have a Twitter account, to see tweets. Information about seminars and workshops is also available through this blog.

Q: My VelevetOptimiser job failed with the following error message: Minimum specified coverage cutoff is higher than the expected coverage. Please choose a minimum coverage cutoff smaller than and re-run. What does it mean?
A: Your read coverage is not sufficient for assembly. Make sure you used correct parameters during trimming.

Q: Can I download BAM index files from Galaxy?
A: Yes. BAM index files, .bai, can be downloaded through Save icon: left click on the icon provides download links for BAM and index files. To copy URLs, use right clicks on the download links. For example, right click over Download dataset open menu where you can copy URL corresponding to BAM file via Copy Link.

Q: Can I use Galaxy-qld for training / workshops?
A: We recommend Galaxy-tut for training activities. If you want to run a workshop on Galaxy-qld, please notify us, so we keep an eye on the server during the training, as big groups can cause spikes in load. Instruct the workshop participants to delete all datasets after the training. This step can be included into a training program. Make sure the participants follow the good user practice for Galaxy-qld.

Q: Why HTSeq produces an empty output?
A: HTSeq uses dataset names for column names and it cannot handle spaces in names. Rename the BAM files (e.g., Tophat on data 1: accepted_hits) produced by tophat to something like c1 or c_1.

Q: Why some tools (Bowtie, Tophat, BWA, BWA-MEM) do not see my FASTQ files?
A: Check the datatype of your FASTQ files. Many tools recognises FASTQ files with the specified quality score offset (datatypes fastqsanger or fastqillumina) but Galaxy assigns fastq datatype to FASTQ files during upload. The appropriate datatype can be specified during upload or by changing datatype in Attributes (the pencil icon) or by FASTQ Groomer tool. If you use FASTQ Groomer please delete and purge the original file.

Q: I am getting errors with Trinity Read Normalisation.
A: The Read Normalisation tool does not like spaces in read names, and many public datasets have spaces in FASTQ read names. Replace spaces in read names with any ‘neutral’ character, such as underscore ( _ ) or convert FASTQ to tabular format and split the names into several columns. Recreate FASTQ from tabular using only one column for read names. Note that Tabular-to-FASTQ assigns fastq datatype to the output. It can be changed to fastqsanger by FASTQ Groomer tool or by changing datatype in Attributes.

Q: My de novo transcript assembly job with Trinitny has failed. Why?
A: Check the log file. If it says ‘Cannot allocate memory’, your job needs more memory. Reduce the size of your input data. Make sure you use Trinity Read Normalisation. Trim the reads and remove short reads.

Q: How can I filter a multi-sample VCF file?
A: Multi-sample VCF files can be filtered on data in Genotype columns using NGS: Annotation > SnpSift Filter. For example, if your samples contain GQ (conditional genotype quality) field, ( GEN[0].GQ < 20 ) will filter all variants with GQ < 20 in the first sample; ( GEN[*].GQ < 20 ) will filter variants with GQ < 20 in any sample.

Q: I am getting errors with GATK tools. Can you help me?
A: Check the error report. If it has something like: “ERROR MESSAGE: SAM/BAM file … is malformed: SAM file doesn’t have any read groups defined in the header. The GATK no longer supports SAM files without read groups“, please add the read groups to your alignment(s) using Picard > AddOrReplaceReadGroups. The read groups can be specified during alignment step with BWA and bowtie.

Q: RNA_START is failing with a custom reference sequence with error:
Fatal error: Matched on [sam_read1] missing header? Abort!
[samopen] no @SQ lines in the header.
[sam_read1] missing header? Abort!
A: Current Galaxy version of RNA_STAR aligner cannot handle short reference sequences. Merge FASTA files to get a bigger reference sequence. If you work with viruses or plasmids, you can merge the viral or plasmid genome with the host genome.

Q: I cannot upload files to Galaxy. Why?
A: If your upload job is queueing for a long time, check your quota usage in the top right corner. If it is over 100%, you cannot upload any data until you free some disk space. If your upload job ended up with an error, check your data. Galaxy recognises files compressed with gzip. It does not like multi-file archives.

Q: My jobs are queueing for long time. What is happen?
A: Refresh your browser by clicking on Galaxy / GVL icon in the top left corner. If your jobs are still in the queue, check Galaxy-qld policy on number of concurrent jobs, to see if you exceeded the limits. If it is the case, your jobs will start after completion of active jobs. For example, if you’ve submitted 10 BWA jobs, 6 should be sent for execusion immediately, and 4 will wait for completion of the active jobs. Start a small test job, such as Select first lines. If the test job is queued for >5 minutes, contact the Galaxy-qld support team.

Q: I deleted several jobs, and now all new jobs stay in queue.
A: Occasionally simultaneous cancellation of several jobs disrupts communication between Galaxy and the computer cluster. Do not cancel jobs in quick succession.

Q: I am from UQ / Australia. I am registered on Galaxy-qld, but my quota is shown as 100 Gb. Can you help?
A: After the registration all new users get the default quota of 100 Gb. All other quotas are assigned manually, and usually it is happen within 24 hours after the registration.