Settings multiple Genome Indices
Hi, I'm working on tool wrapper that needs three reference genome locations (2bit/lastz, all_fasta and novoalign). Is there a way to fetch the three paths from the loc files using a single select box? I've looked around in a couple of tool wrappers on the toolshed, but could find any leads. Something like the following would be ok: Select options from all_fasta.loc -> upon change, set hidden variables to novoalign.loc + lastz_seqs.loc using a filter tag or something against the selected options from all_fasta. Even more simple : access in the command tag like: --novoidx novoalign-loc.options[dbkey=$input.dbkey,path] --fasta all_fasta-loc[dbkey=$input.dbkey,path] etc , using the dbkey from the input file (which is a bam file in this case) In this case, error handling should come from the tool, if one of the indices is missing. I'm currently using three selection boxes, but that looks a but redundant to the user, as they all have to select the same thing. Best, Geert -- Geert Vandeweyer, Ph.D. Department of Medical Genetics University of Antwerp Prins Boudewijnlaan 43 2650 Edegem Belgium Tel: +32 (0)3 275 97 56 E-mail: geert.vandeweyer@ua.ac.be http://ua.ac.be/cognitivegenetics http://www.linkedin.com/in/geertvandeweyer
Hi Geert A simple solution might be: you hand over a string containing all three paths to your wrapper, and then the wrapper splits the string into the three individual paths. Hans-Rudolf On 07/27/2014 04:44 PM, Geert Vandeweyer wrote:
Hi,
I'm working on tool wrapper that needs three reference genome locations (2bit/lastz, all_fasta and novoalign). Is there a way to fetch the three paths from the loc files using a single select box? I've looked around in a couple of tool wrappers on the toolshed, but could find any leads. Something like the following would be ok:
Select options from all_fasta.loc -> upon change, set hidden variables to novoalign.loc + lastz_seqs.loc using a filter tag or something against the selected options from all_fasta.
Even more simple : access in the command tag like: --novoidx novoalign-loc.options[dbkey=$input.dbkey,path] --fasta all_fasta-loc[dbkey=$input.dbkey,path] etc , using the dbkey from the input file (which is a bam file in this case)
In this case, error handling should come from the tool, if one of the indices is missing.
I'm currently using three selection boxes, but that looks a but redundant to the user, as they all have to select the same thing.
Best,
Geert
If I understand you correctly, would this mean generating a custom .loc file for this tool containing the concatenated paths? Or can I concatenate paths from different .loc files somehow, which would allow me to use the default loc files? Geert On 07/28/2014 09:33 AM, Hans-Rudolf Hotz wrote:
Hi Geert
A simple solution might be: you hand over a string containing all three paths to your wrapper, and then the wrapper splits the string into the three individual paths.
Hans-Rudolf
On 07/27/2014 04:44 PM, Geert Vandeweyer wrote:
Hi,
I'm working on tool wrapper that needs three reference genome locations (2bit/lastz, all_fasta and novoalign). Is there a way to fetch the three paths from the loc files using a single select box? I've looked around in a couple of tool wrappers on the toolshed, but could find any leads. Something like the following would be ok:
Select options from all_fasta.loc -> upon change, set hidden variables to novoalign.loc + lastz_seqs.loc using a filter tag or something against the selected options from all_fasta.
Even more simple : access in the command tag like: --novoidx novoalign-loc.options[dbkey=$input.dbkey,path] --fasta all_fasta-loc[dbkey=$input.dbkey,path] etc , using the dbkey from the input file (which is a bam file in this case)
In this case, error handling should come from the tool, if one of the indices is missing.
I'm currently using three selection boxes, but that looks a but redundant to the user, as they all have to select the same thing.
Best,
Geert
-- Geert Vandeweyer, Ph.D. Department of Medical Genetics University of Antwerp Prins Boudewijnlaan 43 2650 Edegem Belgium Tel: +32 (0)3 275 97 56 E-mail: geert.vandeweyer@ua.ac.be http://ua.ac.be/cognitivegenetics http://www.linkedin.com/in/geertvandeweyer
On 07/28/2014 10:02 AM, Geert Vandeweyer wrote:
If I understand you correctly, would this mean generating a custom .loc file for this tool containing the concatenated paths? Or can I concatenate paths from different .loc files somehow, which would allow me to use the default loc files?
This would require a custom .loc file Hans-Rudolf
Geert
On 07/28/2014 09:33 AM, Hans-Rudolf Hotz wrote:
Hi Geert
A simple solution might be: you hand over a string containing all three paths to your wrapper, and then the wrapper splits the string into the three individual paths.
Hans-Rudolf
On 07/27/2014 04:44 PM, Geert Vandeweyer wrote:
Hi,
I'm working on tool wrapper that needs three reference genome locations (2bit/lastz, all_fasta and novoalign). Is there a way to fetch the three paths from the loc files using a single select box? I've looked around in a couple of tool wrappers on the toolshed, but could find any leads. Something like the following would be ok:
Select options from all_fasta.loc -> upon change, set hidden variables to novoalign.loc + lastz_seqs.loc using a filter tag or something against the selected options from all_fasta.
Even more simple : access in the command tag like: --novoidx novoalign-loc.options[dbkey=$input.dbkey,path] --fasta all_fasta-loc[dbkey=$input.dbkey,path] etc , using the dbkey from the input file (which is a bam file in this case)
In this case, error handling should come from the tool, if one of the indices is missing.
I'm currently using three selection boxes, but that looks a but redundant to the user, as they all have to select the same thing.
Best,
Geert
There's an as-yet undocumented method that you can use to get arbitrary fields in a loc file: __get_data_table_entry__(<table_name>, <key_field>, <field_value>, <value_field>) So for the novoalign entry you could do: --novoidx ${__get_data_table_entry__(‘novoalign’, ‘dbkey’, $input.dbkey, ‘path’)} Finally, I like your indexing syntax, and it would be great to use something like that in place of the explicit method call above. Best, J. -- Jeremy Goecks Assistant Professor of Computational Biology George Washington University On Jul 27, 2014, at 10:44 AM, Geert Vandeweyer <geert.vandeweyer2@uantwerpen.be> wrote:
Hi,
I'm working on tool wrapper that needs three reference genome locations (2bit/lastz, all_fasta and novoalign). Is there a way to fetch the three paths from the loc files using a single select box? I've looked around in a couple of tool wrappers on the toolshed, but could find any leads. Something like the following would be ok:
Select options from all_fasta.loc -> upon change, set hidden variables to novoalign.loc + lastz_seqs.loc using a filter tag or something against the selected options from all_fasta.
Even more simple : access in the command tag like: --novoidx novoalign-loc.options[dbkey=$input.dbkey,path] --fasta all_fasta-loc[dbkey=$input.dbkey,path] etc , using the dbkey from the input file (which is a bam file in this case)
In this case, error handling should come from the tool, if one of the indices is missing.
I'm currently using three selection boxes, but that looks a but redundant to the user, as they all have to select the same thing.
Best,
Geert
--
Geert Vandeweyer, Ph.D. Department of Medical Genetics University of Antwerp Prins Boudewijnlaan 43 2650 Edegem Belgium Tel: +32 (0)3 275 97 56 E-mail: geert.vandeweyer@ua.ac.be http://ua.ac.be/cognitivegenetics http://www.linkedin.com/in/geertvandeweyer
___________________________________________________________ Please keep all replies on the list by using "reply all" in your mail client. To manage your subscriptions to this and other Galaxy lists, please use the interface at: http://lists.bx.psu.edu/
To search Galaxy mailing lists use the unified search at: http://galaxyproject.org/search/mailinglists/
Excellent, thanks ! Geert On 07/28/2014 05:02 PM, Jeremy Goecks wrote:
There's an as-yet undocumented method that you can use to get arbitrary fields in a loc file:
__get_data_table_entry__(<table_name>, <key_field>, <field_value>, <value_field>)
So for the novoalign entry you could do:
--novoidx ${__get_data_table_entry__(‘novoalign’, ‘dbkey’, $input.dbkey, ‘path’)}
Finally, I like your indexing syntax, and it would be great to use something like that in place of the explicit method call above.
Best, J.
-- Jeremy Goecks Assistant Professor of Computational Biology George Washington University
On Jul 27, 2014, at 10:44 AM, Geert Vandeweyer <geert.vandeweyer2@uantwerpen.be <mailto:geert.vandeweyer2@uantwerpen.be>> wrote:
Hi,
I'm working on tool wrapper that needs three reference genome locations (2bit/lastz, all_fasta and novoalign). Is there a way to fetch the three paths from the loc files using a single select box? I've looked around in a couple of tool wrappers on the toolshed, but could find any leads. Something like the following would be ok:
Select options from all_fasta.loc -> upon change, set hidden variables to novoalign.loc + lastz_seqs.loc using a filter tag or something against the selected options from all_fasta.
Even more simple : access in the command tag like: --novoidx novoalign-loc.options[dbkey=$input.dbkey,path] --fasta all_fasta-loc[dbkey=$input.dbkey,path] etc , using the dbkey from the input file (which is a bam file in this case)
In this case, error handling should come from the tool, if one of the indices is missing.
I'm currently using three selection boxes, but that looks a but redundant to the user, as they all have to select the same thing.
Best,
Geert
--
Geert Vandeweyer, Ph.D. Department of Medical Genetics University of Antwerp Prins Boudewijnlaan 43 2650 Edegem Belgium Tel: +32 (0)3 275 97 56 E-mail: geert.vandeweyer@ua.ac.be <mailto:geert.vandeweyer@ua.ac.be> http://ua.ac.be/cognitivegenetics http://www.linkedin.com/in/geertvandeweyer
___________________________________________________________ Please keep all replies on the list by using "reply all" in your mail client. To manage your subscriptions to this and other Galaxy lists, please use the interface at: http://lists.bx.psu.edu/
To search Galaxy mailing lists use the unified search at: http://galaxyproject.org/search/mailinglists/
-- Geert Vandeweyer, Ph.D. Department of Medical Genetics University of Antwerp Prins Boudewijnlaan 43 2650 Edegem Belgium Tel: +32 (0)3 275 97 56 E-mail: geert.vandeweyer@ua.ac.be http://ua.ac.be/cognitivegenetics http://www.linkedin.com/in/geertvandeweyer
participants (3)
-
Geert Vandeweyer -
Hans-Rudolf Hotz -
Jeremy Goecks